THE MELLON PROJECT

MUSEUM EDUCATIONAL SITE LICENSING PROJECT:
Studying the Economics of
Network Access to Visual Information

 

INTERIM DRAFT REPORT

THE ECONOMICS OF NETWORKED INFORMATION:
Theory, Methods and Preliminary Results

August, 1997

 

 

 

 

 

 

 

 

Howard Besser

School of Information Systems and Management
University of California, Berkeley

Robert Yamashita

Science, Technology and Culture
Liberal Studies
California State University, San Marcos
&
School of Information Systems and Management
University of California, Berkeley

TABLE OF CONTENTS

EXECUTIVE SUMMARY *

Background *

Overview of Cost Center in the Digital Model *

Analog Model Overview *

Provisional Summary *

INTRODUCTION *

ECONOMIC FRAMING *

MESL – The Digital Distribution of Images *

Internal Cost Centers *

Production Environment *

Processing Environment *

Deployment Environment *

Security *

Usage Environment *

Technology Infrastructure *

Social Organization *

SUMMARY *

REFERENCES *

APPENDICES *

Appendix 1 – METHODS *

Appendix 2 – BACKGROUND *

Appendix 3 – THE ANALOG DISTRIBUTION CHAIN *

NOTES *

EXECUTIVE SUMMARY

This document is a preliminary report of our Mellon-sponsored investigation of the costs and uses associated with networked distribution of visual information for educational purposes. This study focuses on an existing collaboration between seven museums and seven universities to distribute digital images and associated text as part of the Museum Educational Site Licensing (MESL) Project.* It compares these costs with those associated with analog image distribution methods. The grant period for the Mellon study extends for 18 months from September 1, 1996 through March 1, 1998, and this report is being prepared at the 12-month mark of the study. It is very tentative and preliminary, as data collection for this study did not begin until the spring of 1997.

The fact that we focused our interest on MESL is both enlightening and limiting. Advantages include that this is the first attempt at large-scale distribution of digital images to the educational community, that we can study seven heterogeneous cultural repositories contributing images to the project, and that we can examine seven different university environments delivering images and text to their end-users. Limitations include the difficulty posed by comparing environments that are sometimes so different that we can’t even find common units of measurement, as well as the fact that MESL chose only one of many possible distribution/delivery models – making it difficult to extrapolate from MESL findings into other models.

We nonetheless want to make the findings of this study broadly applicable. We therefore chose to identify the primary elements of the digital image delivery pathway that runs from the original object to the end user’s access of its digital representation. We focus our attention on the cost centers that are likely to be present in the delivery pathway of most distribution models. By identifying and isolating these cost centers, we believe that we can provide an important framework for future projects, even those using models where any given cost center may move into a different type of organization than in the MESL model. The exercise allows a systematic comparison with other forms of image distribution such as those found in analog slide libraries.

The economic examination of the digital distribution methods under MESL is still in process. To date, we have identified the basic cost centers and summarized the activities and functions that take place at each. We have collected the primary source data for these centers. Many of the technical hurdles have been addressed including defining methods to manage problems with data quality and comparability. The next step will be systematic intra- and inter-organizational analyses to identify the critical pathways and costs at each of these centers. This requires summarizing resources such as labor and direct capital investments and other indirect costs associated with the specific activities in each center. Indeterminate costs such as general infrastructure requirements and prior experience will be summarized. We will attempt to note the types of activities that are enabled at certain thresholds of resource allocation. And we will focus on information that will be most relevant to future implementers trying to determine resources needed to provide particular forms of functionality. The analysis is difficult not only because of MESL’s innovative nature but also because of its formal design. The many-to-many experiment with varying institutional characteristics and different data collection points makes simple mechanical comparisons problematic.

The attached draft report presents the project’s framing, methods and preliminary findings of the digital image distribution methods under MESL. It identifies the major structural relationships, central practice environments, and delivery and usage practices in the production and distribution of digital images. It provides a tentative identification of the key cost centers within these environments and outlines some provisional "critical paths" required for the effective creation and distribution of digital images and major cost indicators. The appendix to this document includes a discussion of methods, background concepts, and a brief overview of the analog distribution chain. The final report will include a more detailed account and better summaries of costs and resources needed for each of the cost centers, and an examination of the costs of operating an analog slide library and comparison with the cost centers along the digital distribution and delivery pathway.

Background

The digital image delivery pathway contains five practice environments that capture the basic work processes. They include (1) production of digital images and their associated text, (2) data processing, (3) database deployment, (4) security, and (5) user access. Figure 1 outlines the basic relationships in the delivery pathway.

Figure 1. The Distribution Pathway

The production environments include making digital representations of object and their associated text. The processing environment takes the data files and places them into a sequential database container. The deployment environment takes this container and creates a usable database, attaches a search engine, retrieval frames, and interface. Additional value can come through a re-packaging of data and enhancing its value by providing groupings and conceptual views. The security environment restricts access to these data. Finally, the usage environment impacts how users interact with the data. These environments loosely mirror the social organization in the formal implementation of the MESL project (see Appendix 2 – BACKGROUND).

The pathway is framed by two sets of constraints. The terms and conditions frame the kinds of permission that both enable the digitization of images and limits how they are used. This primarily impacts the production environment and conditions the kinds of security apparatus that gets implemented. The preference and relevance frame determines the kinds of interest in these images. This ultimately determines what gets selected. Deciphering the impact of these conditional frames on the delivery pathway is critical for understanding the overall trajectory of the costs and usage of a digital distribution system. Ignoring any specific element along this pathway will yield a misunderstanding of both the costs and the benefits of the system.

Overview of Cost Center in the Digital Model

Within the distribution pathway, a number of distinct cost centers can be derived for each of the practice environments. A cost center constitutes a collection of linked activities that appear to be required to accomplish a particular task. For example, the process of identifying the legal rights to digitize and license an object would be a necessary cost center for an image producer. This cost center would be different from the process required to identify images that consumers might want to use, or the center that involves the actual digitization of the object. There are different kinds of relationships between cost centers within environments. Some are procedural with linear dependencies (e.g. one step requires another), others operate as parallel processes (e.g. they happen simultaneously), and some are discrete either-or operations. Figure 2 outlines the provisional cost centers for the distribution model.

Figure 2. Digital Distribution Model

Each cost center consists of a set of discrete procedures. In principle, cost figures can be attached to each step. Costs are likely to fluctuate radically over time, so we view our primary contribution as identifying cost centers and associated functionality. We will attempt to provide more long-lasting methods for examining each of these cost centers, such as factoring in variables such as learning curves. Our study will also identify the human knowledge and background and organizational infrastructures needed in order to complete a project such as this. A key part of our effort is to understand what it took to create, distribute, and mount the digitized materials used in the MESL project

Analog Model Overview

A comparative examination of analog image distribution/delivery systems (slide libraries) is in its opening stages. The analog model is an important comparative case because historically slide libraries have been the primary method for the mass distribution of cultural heritage information to the educational community. We have grouped cost centers in ways to expedite comparisons between analog and digital versions of each.

We expect to find that a number of the cost centers for a slide library closely duplicate efforts already undertaken by museums, and that these costs might be greatly diminished in a more cooperative environment. For example, the slide library’s cost center for textual description of the object depicted in a slide might be replaced by a payment to a museum for conversion of their own textual description of that object into a form acceptable to the slide library. This study will provide significant comparative evidence for the overall effectiveness, strengths and weaknesses, in the digital image distribution model.

Provisional Summary

We feel that this is a propitious time for this study, as two museum consortia are currently developing plans to distribute digital images to the museum community. Both the Art Museum Directors Association’s AMICO project and the American Association of Museums’ Museum Licensing Consortium are currently designing both their distribution/delivery schemes and their business plans. Both could benefit from a better understanding of the cost centers and efforts involved in making digital images and accompanying text available to the educational community. Thus far, these consortia have focused their attention on framing issues (such as terms and conditions of use) and not the core mechanics of accomplishing their objective. Their work on models for production, processing, and deployment, as well as development of cost models can be greatly informed by the results of this Mellon-sponsored study.

To date the evidence suggests that the digital distribution model should not be viewed as a direct replacement of analog distribution systems. The lack of substantive infrastructure in the academic environment (e.g. digital classrooms, appropriate workstations and labs) and useful tools to help manipulate image sets currently makes the functional use of these system problematic. In this study we will identify these barriers to widespread adoption of digital distribution/delivery systems, as well as advantages that have enticed faculty and students to use these systems despite their drawbacks. But for the foreseeable future, the digital distribution model should be understood as a supplement to, rather than a replacement for analog distribution systems for cultural heritage information.

We also expect that a study of value assigned to digital images by users will be relatively low. We surmise that museums should not expect an income stream from sales of images in this format, and (from a monetary standpoint) should at best hope for cost-recovery. But preliminary anecdotal evidence from the MESL project suggests that digital distribution can yield enticing benefits not related to a potential income stream. This might take forms as diverse as: increasing the museum’s public profile and attracting new use and attention; in promoting new capabilities, work procedures, and staff relationships within the museum itself; encouraging standard practices and cooperative ventures between museums.

The Mellon study will be completed in early 1998.

THE MELLON PROJECT

MUSEUM EDUCATIONAL SITE LICENSING PROJECT:
Studying the Economics of
Network Access to Visual Information

 

INTERIM DRAFT REPORT

THE ECONOMICS OF NETWORKED INFORMATION:
Theory, Methods and Preliminary Results

August, 1997

 

 

 

 

 

 

 

 

Howard Besser

School of Information Systems and Management
University of California, Berkeley

Robert Yamashita

Science, Technology and Culture
Liberal Studies
California State University, San Marcos
&
School of Information Systems and Management
University of California, Berkeley

INTRODUCTION+

This document is a preliminary report of our Mellon-sponsored investigation of the costs and uses associated with networked distribution of visual information for educational purposes. This study focuses on an existing collaboration between seven museums and seven universities to distribute digital images and associated text as part of the Museum Educational Site Licensing (MESL) Project. The grant period for the Mellon study extends for 18 months from September 1, 1996 through March 1, 1998, and this report is being prepared at the 12-month mark of the study. It is very tentative and preliminary, as data collection for this study did not begin until Spring of 1997.

MESL was designed as a prototype demonstration of using "digital imaging and network technologies" to "make cultural heritage information more broadly available." It consisted of two basic objectives: (1) to develop, test and evaluate procedures and mechanisms for the collection and dissemination of museum images and information, and (2) propose a framework for a broadly-based system for the distribution of museum images and information on an on-going basis to the academic community. (MESL: Goals and Objectives, 2/22/95:1-2) The MESL project provides us with a unique opportunity to examine costs and uses of digital images delivered over campus networks.

The fact that we are studying a single project (MESL) is both enlightening and limiting. Advantages include that this is the first attempt at large-scale distribution of digital images to the educational community, that we can study 7 heterogeneous cultural repositories contributing images to the project, and that we can examine 7 different university environments delivering images and text to their end-users. Limitations include the difficulty posed by comparing environments that are sometimes so different that we can’t even find common units of measurement, as well as the fact that MESL chose only one of many possible distribution/delivery models -- making it difficult to extrapolate from MESL findings into other models.

We want to make the findings of this study as broadly applicable as possible. We have therefore chosen to focus in on cost centers that are likely to be present as steps of the distribution chain (from the original object until its digital representation reaches the hands of the end user) in most potential distribution/delivery models. By identifying and isolating these cost centers, we believe that we can provide an important framework for future projects, even those using models where any given cost center may move into a different type of organization than in the MESL model.

For each cost center at each step along the chain, we will identify estimated cost figures. Costs are likely to fluctuate radically over time, so we view our primary contribution as identifying cost centers. We will also attempt to provide more long-lasting methods for examining each of these cost centers, factoring in variables such as learning curves. Our study will also identify the human knowledge and background and organizational infrastructures needed in order to complete a project such as this. A key part of our effort is to understand what it took to create, distribute, and mount the digitized materials used in the MESL project

We feel that this is a propitious time for this study, as two museum consortia are currently developing plans to distribute digital images to the museum community. Both the Art Museum Directors Association’s AMICO project and the American Association of Museums’ Museum Licensing Consortium are currently designing their distribution and delivery schemes. Their business plans could benefit from a better understanding of the cost centers and efforts involved in making digital images and accompanying text available to the educational community. Thus far these consortia have focused their attention on framing issues (such as terms and conditions of use). Their work on models for production, processing, and deployment, as well as development of cost models can be greatly informed by the results of this Mellon-sponsored study.

To date, we have identified the basic cost centers and summarized the activities and functions that take place at each. The next step will be an analysis of the data that has already been gathered, identifying costs at each of these centers. This requires summarizing resources needed and other costs associated with the specific activities in each cost center. Costs such as labor, capital, and infrastructure, as well as prior experience will be summarized. We will attempt to note the types of activities that are enabled at certain thresholds of resource allocation. And we will focus on information that will be most relevant to future implementers trying to determine resources needed to provide particular forms of functionality.

This draft interim report presents the project’s framing, methods and preliminary findings of the digital image distribution methods under MESL. It identifies the major structural relationships, central practice environments, and delivery and usage practices in the production and distribution of digital images. It provides a tentative identification of the key cost centers within these environments and outlines some provisional "critical paths" required for the effective creation and distribution of digital images and major cost indicators. The appendix to this document includes a discussion of methods, background concepts, and a brief overview of the analog distribution chain. The final report will include a more detailed account and better summaries of costs and resources needed for each of the cost centers, and an examination of the costs of operating an analog slide library and comparison with the cost centers along the digital distribution and delivery pathway.

ECONOMIC FRAMING

The economic evaluation has to recognize that MESL constitutes an experiment in digital entity production and distribution. The MESL model was developed under a set of limitations and constraints and needs to be understood as only one of many possible models. Because of its experimental character, this model is more than likely not to be replicated in production environments. The analysis therefore has to move beyond the specific institutional structure of MESL to an examination of the necessary functions to distribute digital entities.

MESL’s structural organization highlights the segmentation of processes that form a distribution pathway in the delivery of digital images. The segmentation groups fundamental operations that any digital image delivery model has to address. There are physically distinct zones: the production of images and associated text, processing of image and text data and the creation of functional digital entities, the deployment of entities through a distribution system, security to control access, and usage of these entities by enduser. Each zone constitutes an "environment" that locates technical activity required for digital image delivery. They can be described as:

Collectively these environments form a distribution pathway. This basic schema allows for the characterization of the functional digital image delivery pathway (Figure 1).

Figure 1. Basic Digital Image Distribution Model

While the structural order mirrors that of MESL, the functional component elements for the economic analyses are different in several respects. Under MESL each environment was institutionally discrete – museums produced, universities deployed. In other models responsibility for processes can reside internally to an institution or they can be shared or contracted out among several institutional sites. However, as in MESL, the terms and conditions of use, and the preference and sense of relevance of endusers will frame activity within each zone. Organizationally, the functional model first identifies a processing environment. This environment includes both MESL’s central processing facility and a portion of the work done by universities – the integration data elements into functional digital entities, and the aggregation of the multiple data sets into a cohesive sequential database container. Second the functional model adds a formal security environment because the implementation of any security apparatus tends to engage actions that are distinct from deployment or use. More importantly, during the later stages of the MESL, several institutions began to move away from simple access controls to more formal security implementations (e.g. authentication). These change the costs of the infrastructure requirements.

Each practice environment encapsulates a production process that includes multiple component tasks. The task of the economic evaluation is to locate all these elements and determine their relationship and relevance. Identification requires evaluating their economic impact since the cost of these tasks range from continuous direct charges to one-time charges, indirect assessments, or simply technical concerns.

MESL – The Digital Distribution of Images

The examination of the practice environments can be broken out into three distinct areas that define the economics of the digital distribution of images: formal cost centers, technological infrastructure, and the institutional organization. Each production environment consists of a number of cost centers. A cost center constitutes a collection of linked activities that appear to be required to accomplish a particular task. For example, the process of identifying the legal rights to digitize and license an object would be a necessary cost center for an image producer. This cost center would be different from the process required to identify images that consumers might want to use, or the center that involves the actual digitization of the object. There are different kinds of relationships between cost centers within environments. Some are procedural with linear dependencies (e.g. one step requires another), others operate as parallel processes (e.g. they happen simultaneously), and some are discrete either-or operations. The technological infrastructure constitutes the ambiguous connections between practice environments or between clusters of specific activity. For example, the delivery system between the production environment and processing environment could be the Internet or a mail/package delivery system. Similarly, the connection between the selection of an object to digitize and the digitization process could be personal knowledge, notation on a card or in a database, a phone call or electronic mail. Institutional organization is the defining social structure that impacts and shapes practice. This includes the distinct participation of operational units and sponsorship. For example, in both museums and universities, the department that housed primary responsibility shaped how the project operated. The functional relationship can also be top-down or bottom-up. In most cases, this clearly determines the commitment of actual resources rather than simple "in-king" contributions.

Internal Cost Centers

A number of discrete cost centers can be identified for each practice environment. A cost center encapsulates a set of processes and activities and result in a distinct end product. Economic evaluation requires identifying and measuring the resource commitment to accomplish the task and includes calculating the costs of machines, mechanical process, skilled personnel, and the necessary infrastructure. Figure 2 provisionally outlines these cost centers.

Figure 2. Provisional Outline of Key Cost Centers

Analyses of these cost centers for MESL are difficult. The technical reports ask for information on automation and personnel requirements. These resources form the key data for determining hard economic costs. A review of the data in these reports demonstrates wide discrepancies in the number of persons (or person hours) required accomplishing the same task. It could be that the specific differences are an artifact of institutional culture (it is done that way), data quality, or projection of the number of bodies in a given unit. Assessment of the numbers therefore needs to begin with an internal examination of the institution, before making comparisons across sites. Finally, specific questions as to the infrastructure requirements were not asked. In many instances it can be assumed that there are some minimal set of technical elements need to be in place before a task can occur. The different level of existing capabilities can result in different qualities in results. Nonetheless, their physical impact on the overall cost picture needs to be documented. Specific assessment will come through selected follow-up interviews and inferred across sites. What follows is a preliminary overview and summary of data for the key cost centers in each practice environment.

Production Environment

The production environment creates raw digital images and their associated text. The specific requirements for these images and text can be determined by third parties. Under MESL, the primary producers of digital images and text were officially restricted to museums. However, each university reported on having to produce their own set of additional digital images and texts to fulfill the specific interests and course needs of local faculty. The cost centers in the production environment include selection, digitization, documentation, data conversion, and object transfer (summarized by Environment Table 1).

The production environment was examined in Section 2 of the technical report. This environment includes a number of distinct activities – selection, permission, digital imaging, text data selection, and transmission. An additional activity, not identified in the museum technical report, was the effort required for fixing inadequate images or data files. Each activity within the environment constitutes a cost center.

Environment Table 1. Production

Outline of Cost Centers and Primary Data Source

Cost Center

Description

Source

  1. Selection

Identification of objects for digitization. This includes collection specifications and Rights Clearance

MTR sections: 2.1.2, 2.2.3, 3.2

  • Digitization
  • The actual creation of a digitized object – requires the digitization of either the object or a representation. Usually consists of two processes. First an Analog photograph of the object is taken and then this photograph is digitally captured.

    MTR sections: 2.3.1, 3.3.

  • Image Data Conversion
  • Insure that the digital images correspond to the required specification.

    MTR, OR: Michigan

  • Documentation
  • This is the extraction of existing collections and curatorial information about the objects from local information systems. It also requires extracting these data from these databases and placing them in a uniform format that fits the proposed classification scheme.

    MTR sections

    2.3.2, 3.4

  • Text Data Conversion
  • Insure that the text data fields content and labels correspond to required specifications.

    MTR

  • Error Correction
  • Working with the processing facility to rectify errors in transmitted data files. May require re-accessing the local information systems to re-digitizing the image.

    OR: Michigan

  • Object Transfer
  • This is the physical sending a set of digital images and an accompanying delimited file consisting of the text documentation. It requires some form of data file compilation, placing it some transportable form (e.g. tape or CD-ROM) and then physically shipping the data (electronically or via snail mail post office or package delivery service).

    MTR sections

    2.3.3, 3.5

    Notes: MTR = museum technical reports, UTR = university technical reports, OR = other reports,
    FGT = focus group transcripts, LS = list server communication, MPR = museum proposals,
    UPR = university proposals

    One of the important tasks is to decipher the range of possible objects available for imaging and the processes used to select objects (including permission to digitize them) and then process them. Image selection and gathering permissions present the starting point for establishing the cost of the MESL process. Section 2.1 of the technical report described the collection, and Section 2.2 collected data on the selection and permission processing. Image selection is reported as a matter of fact. It either occurred through a solicitation of interest by participant faculty, or simply mandated as part of the museums’ provided data. In order to establish the significance of specific costs there will need to be an exploration of the use of these different kinds of selection criterion for the various sets of images.

    Assessing actual digitization cost under MESL is difficult. While MESL image producers offered over 9000 images in two distributions, the vast majority of digital images were captured prior to MESL(summarized by Table 1).

    Table 1. Number of Images Delivered

    Institution

    Distribution

    (section 2.3.x)

    processing

    (section 2.3.1.1)

     

    R. 1

    R. 2

    Total

    Already

    New

    FMCH

    584

    765

    1349

    701

    648

    GEH

    500

    500

    1000

    0

    1000

    HUAM

    527

    509

    1036

    875

    161

    MFAH

    496

    515

    1011

    0

    1011

    LOC

    1400

    750

    2150

    2150

    0

    NGA

    739

    825

    1564

    1548

    16

    NMAA

    541

    459

    1000

    1000

    0

    Total

    6275

    2835

     

    54%

    46%

     

    69%

    31%

    Critical assessments of cost need to separate out what are real costs (specific for the project) or what reflect "recapture" of previous expenditures – collateral benefit. Such evaluation needs to extend beyond the simple notion of having digital images for an external audience. Other internal units could benefit and use the images for new purposes (automating and integrating registrar, collections, and curatorial information etc.) and could lead to new internal digitization efforts.

    Critical assessment also needs to account not only for the physical costs of digital imaging but also what actually gets imaged – the object directly using a digital camera, or an analog slide. If an analog slide is used, questions of costs associated with the photography need to be addressed. Experience with digitization also needs to be explored. The theoretical costs of the first pass are always higher than subsequent efforts. At museum sites efforts to implement MESL image production ranged from well developed internal procedures to out-sourcing of work. Another technical issue is whether the existing digital image was offered "as is" or whether a derivative generated from it was supplied. While the "value" of the derivative would be lower than the original image, there would be an additional cost incurred in creating the derivative (one that is probably acceptable). The experience with digital imaging project appears to underlie choices, but the perceived use-value attached to those images clearly has an impact. Simple expedience to deliver digital files led to out-sourcing (GEH, MFAH, LOC).

    Another final set of issues is to define the costs of producing accompanying text documentation. The MESL technical report addressed the cost incurred for other data such as structured text derived from the collections systems or unstructured data from other information systems. The most significant, and probably the most costly, technical issue focuses on taking existing digital versions of text documentation and putting it in a form where it could be used by MESL. This apparent simple conversion process can pose some interesting and complex problems. Most collection data exist in legacy systems and their extraction into a useable shared form is not necessarily a simply and mundane transaction. The substantive question of translating the local classification vocabularies clearly had some impact on the quality and quantity of the associated text. The cost of the technical implementation of merging these different data systems in the production environment needs exploration.

    Processing Environment

    The processing environment is the creation of a sequential database container containing the accessible digital entities and includes quality control, aggregation, and integration (see Environment Table 2). The relationship between cost centers tends to procedural. Under most distribution models, individual producers will provide initial sets of digital images and text information. The individual data files will need to be processed to insure for quality and structural standardization. These files will then have to be integrated into a larger data container that will include files from other providers.

    Environment Table 2. Processing

    Outline of Cost Centers and Primary Data Source

    Cost Center

    Description

    Source

    1. Quality Control

    Object Checking, documentation checking, and data standardization (bit checking)

    OR – Michigan cost report.

  • Aggregation
  • Entity creation: the merging of image and text data to create a sequential data file consisting of the all the data elements for each individual site. This primarily requires the parsing of the data units and their organization into distinct data structures.

    UTR sections 2.1.3

  • Integration
  • Cross-Site Entity Linkage which requires insuring that the structure of data elements from the sequential data files match. The linking of these data files into a single collection constitutes the basic sequential database container.

    UTR sections 2.1.3

  • Transfer:
  • The sequential database container is made accessible for formal database development and deployment.

    UTR

    Notes: MTR = museum technical reports, UTR = university technical reports, OR = other reports, FGT = focus group transcripts, LS = list server communication, MPR = museum proposals, UPR = university proposals

    Under MESL this environment was divided between the central processing facility at the University of Michigan and the individual university deployment sites. Future distribution schemes could move more elements into the central distribution site where the work would only be done once. However, certain procedures such as image derivation and lossy compression requires an agreed upon standards for image size(s), qualities, and compression ratios (see Besser & Stephenson 1996).

    The processing environment for MESL was located at the University of Michigan. The activities of the site included basic quality control, data standardization, and the delivery of objects and text to the deployment sites. During each of MESL’s two main distributions, the central distribution site took batches of images and text received from each museum and forwarded these on to each university. The deployment sites need to create digital entities through an aggregation of the data received from individual providers. They then needed to integrate these entities into their database so they could be manipulated. Additional processing was required at the university deployment sites. Raw data files needed to parsed into a single database container, and then aggregated with other digital objects.

    In terms of assessing the cost of insuring data quality, the University of Michigan submitted a cost report on its activities on its distribution activities. The report compared costs over the two distributions. It showed a sharp reduction in overall costs (over 50%) due to having processes in place for the second distribution. Table 2 summarizes these data.

    Table 2. Error Checking Costs

    Distribution

    N

    $

    ave.

    1

    4789

    $ 32,498

    $ 6.79

    2

    4325

    $ 13,938

    $ 3.22

    Total

    9110

    $ 46,436

    $ 5.10

    While average cost for MESL checking was $5.10 per image (direct costs), with the second distribution averaging $3.22. Subsequent deliveries would probably only see a slight decline. What is important is to distinguish between the types of errors incurred and their source. Its important to distinguish error related to corrupt individual files, errors created by operators ("human error"), and errors that are a function of merging data from different information systems. Another issue is to determine what the other side of the cost estimates at "fixing" errors provided in the report that does not reflect the costs incurred by individual museums. A final issue is to distinguish between what was checked by the central-processing site and what additional efforts had to be done at each university in order to make the collected data base usable.

    The universities took museum data files from the central processing facility and created the basic sequential database container. With text data, they took the flat delimited text files and parsed them using a variety of application tools (Perl scripts, Excel, Filemaker Pro, Microsoft Access). In most cases, the tools both created HTML pages for each record and loaded these data into a database. The different sites used a variety of procedures. For example, instead of "pre-computing" HTML-formatted records separate from the database, Virginia stored all their records in "pseudo" SGML format, ran database queries against this stored data (using Open Text), and generated HTML results pages from it on-the-fly.

    During the first distribution, the universities had significant problems parsing and loading the text data due to: poor constructed or undelimited fields, each museum using a different set of delimiters, line-feeds in the middle of records, and the use of different character sets. Many of these problems disappeared in the second distribution, as the MESL participants as a whole attempted to standardize and better specify delimiters and character sets. The MESL experience made it clear that data specification must be extremely precise, and that a heterogeneous pool of institutions during a pilot study can reveal divergent practices that were not taken into consideration during initial attempts at specification. Over the course of the project, MESL participants appear to have developed a set of standard specifications precise enough to assure consistent syntax, but other follow-up projects will have to tackle the even more difficult problems associated with standardization on data values.

    With images, most universities made user interface and general design decisions based upon particular image sizes/qualities. Instructors expressed concerns over image size including that they be: big enough for classroom projection, be as big as possible yet fit on the "average" screen without scrolling, fit within a specific application without scrolling, etc. Many sites thus had an investment in a particular size of image. Table 3 illustrates how image sizes and qualities varied widely between the different deployment sites, even among well-recognized "sizes" such as thumbnails (although most sites delivered compressed images, a comparison of compression ratios or quality is very difficult, as there is no standard scale upon which to measure this).

    Table 3. Image sizes & formats delivered at each site

    (complied by Christie Stephenson)

    University

    Thumbnail

    Medium

    (Screen Size)

    Large

    Other

    American

    50 x 50

    640 x 400

    off-line

     

    Columbia

    100 x 70 GIF 89

    350 x 250 JPEG

    700 x 500 JPEG

    1200 x 900 JPEG

    Cornell

    120 pixels max

    dimension

    390 pixels max dimension

    as supplied;

    Photo CD images converted to JPEG

     

    Illinois

    125 pixels high

    JPEG/JFIF

    400 pixels high

    JPEG/JFIF

    Compressed but not resized

    JPEG/JFIF

     

    Maryland

    150 pixels max

    dimension

    BMP/GIF

    700 pixels max

    dimension

    BMP/GIF???

       

    Michigan

    90 pixels max dimension

    GIF 89A

    640 pixels max dimension

    JPEG/JFIF

    960 pixels max dimension

    maximum--the full size image supplied by the museum

    Virginia

    130 pixels high

    GIF

    600 pixels high

    JPEG/JFIF

    off-line

     

    None of the university implementations supported on-the-fly derivation of smaller images. When the images were received from the central distribution site each university generated several sizes of derivative images (thumbnail, large image, and often one or more in-between) for delivery. Applications like Debabelizer and ImageMagick make this derivation process relatively simple in an unattended batch mode. Though the batch post-processing worked well in creating most derivative images, there were some substantive technical problems. The most significant had to do with data loss due to file compression. Some images were already lossy-compressed. The amount of lost data was amplified as the image was uncompressed, reduced or resized, and then recompressed again. Particular types of images such as those in a Photo CD format proved quite difficult to work with. Finally, batch compression had problems working across multiple forms of content (e.g. line drawings, engravings, and paintings); future projects should separate line drawings from continuous tone and compress these in separate batches.

    The quality of the supplied images for MESL created another set of substantive problems. Although the universities were generally pleased with the digital images they received, they experienced a number of problems. Some of these problems were clearly due to the scanning process followed by the museums, and any future efforts to distribute digital images should note these. Some universities observed that certain images were of poor quality, probably because they were scanned from poor quality intermediates. Some images delivered were too small for the universities to make effective use of them. Some of the images were dark and muddy, probably because they had either not been color-corrected, or had only been viewed on one particular monitor/platform combination. It is clear that we still lack the proper color management tools to assure that images will look good and consistent from one platform and monitor to another. According to the Columbia technical report, "The quality of the digital images varied from museum to museum, but in general we found the resolution to be too low when compared with images we have been able to obtain commercially." And when Columbia faculty compared projected slides alongside projected digital images of the same object, they found the quality of the digital image sorely lacking. Other problems came from the procedures followed, and may have come from anywhere along the chain leading from the museums to the central distribution site to the universities. Some image files were corrupted, and others were missing, misnamed, or miss-referenced. One final image issue that was handled differently among sites was the placement of borders around images and the matting of backgrounds (particularly on thumbnails) to create images of a consistent aspect ratio.

    Deployment Environment

    The deployment environment results in the production of a functional database system and mounting to a local delivery system. Deployment consists of four fundamental activities: database design, database creation, database tools, and database interface. Within each of these activities there are a number of distinct cost center elements. For example, under database creation there is database element specification and data preparation. Environment Table 3 summarizes these centers.

    Environment Table 3. Distribution

    Outline of Cost Centers and Primary Data Source

    Cost Center

    Description

    Source

    1. Database Design

    Design the basic system: identifying elements of interest and outlining capabilities and functionality.

    LS, UTR

  • Database
  • Create the physical structured database from selected elements and processed data.

    UTR

    • Database elements

    Selection of field to use from database. This is linked to the general concept on what kinds and level of capabilities are desired for the system.

    UTR

    • Data Preparation

    Prepare data for the physical database. Can include converting data elements such as images to meet the required specifications of the database (e.g. thumbnails, resolution).

    UTR

    1. Database tools

    Develop the capabilities to manipulate the database.

    UTR, LS, OR: Berkeley

    • Process identification

    From the database structure identify the kinds of processes that are wanted.

    UTR section 2.1.4, OR

    • Search and Retrieval

    Design and provide functionality for the kinds and types of tools provided to managers and endusers for the manipulation of the data

    UTR section 2.x.x, OR

    1. Database Interface

    Design and build the front end for accessing the database. Develop access methods to the database tools.

    UTR

  • Database Packaging
  • Value-added enhancements to the database can be achieved through the addition of new conceptual views that repackage existing data in new ways.

    UTR

  • Transfer
  • Mounting to an information retrieval system

    UTR

    Notes: MTR = museum technical reports, UTR = university technical reports, OR = other reports, FGT = focus group transcripts, LS = list server communication, MPR = museum proposals, UPR = university proposals

    The university technical report requested information on MESL implementation of the digital distribution system. The development and implementation of this system required several distinct procedural activities ranging from simple acquisition of data storage capacity, to the formal design of the database, and implementation of the delivery system, to the required support of endusers. Each set of activities constitutes a discrete cost center.

    The server implementation at each distribution site varied from fairly robust dedicated servers to modified workstations (or a combination). While all sites received the same data set, some sites reported need to acquire additional storage resources. These substantive differences in hardware infrastructure requirements need to be linked to the formal organization of MESL within the university system. A second infrastructure issue emerged in assessing both the anticipated clients and MESL implementation in the classroom. Many sites were forced to address the formal limits of their infrastructure – even though they all reported having dedicated digital laboratories or classrooms.

    Each university independently designed its own system for deployment of images and text on its campus. Six of the seven sites developed their MESL delivery under web-based strategies. Different operating systems and approaches dictated formal differences in interface design and possible differences in functionality (section 2.1.4). These subtle differences in design based on formal infrastructure choice have a potentially significant cost impact. This relationship needs to be explored in greater detail. Both Illinois and Cornell began with another delivery system then moved onto the Web midway through the project. The seventh site at the University of Maryland used a proprietary system that added greater functionality for pedagogical specific classroom activities. Interestingly, they also had a web version of MESL under development, but that system lacked the enhancements that made the system classroom friendly.

    All of the implementations except Virginia provided a browse function to allow the user to scan through large batches of images and records without first performing a query, but most of the browse applications limited the user to browsing within only a single museum at a time. All of the Web-based delivery systems provided searching through HTML forms that generated cgi-scripted calls to a back-end database/search engine. Back-end databases/search engines included products such as Filemaker Pro, Microsoft SQL server, and Glimpse) and locally designed systems (such as Full Text Lexicographer) (see Table II). The linking of HTML forms-based queries to back-end databases is still very complex, but new applications (such as Cold Fusion) should make this simpler in the future.

    Table 4. Back-end search engines employed at each site

    University

    Search Engine

    American

    Flat database files

    (in-house)

    Columbia

    Glimpse

    Cornell

    Filemaker Pro

    Illinois

    Microsoft SQL Server

    Maryland

    Microsoft Access

    (customized with Visual Basic)

    Michigan

    Full Text Lexicographer

    (locally developed)

    Virginia

    Open Text

    Because each MESL site took such a different approach to mounting the same original set of data, it is difficult to compare and evaluate these in any systematic way. Implementation differences appear to be heavily influenced by factors specific to each site. Two issues appear to standout.

    There was a wide variety in the way the various university implementations looked to a user. A group of Berkeley students preformed the only cross-implementation study, comparing six of the MESL implementations. The findings of their informal study are:

    The central issue of importance is that each site chose to index a different set of fields. Some sites chose to provide keyword access while others did not. Some sites also provided access by categories of local interests (such as by course using the image). Of significance is that in many instances, the "searchable fields" on the user’s query form were really composed of indexes made off of a variety of different fields in the database. The fact that different sites combined their indexes in different ways was one of the factors that periodically led to radically different search results between sites on the same query.

    As part of a comparison between implementations, the Berkeley students were also asked to perform the same search at each site. Because the set of searchable fields presented to the user differed from site to site, students needed to use their own best judgment to replicate the search as closely as possible. These searches yielded vastly different results from site to site:

    Reasons for divergent search results included: sites combining different sets of original data fields into unified indexes, different search engines, and whole-word versus character-string searches on various fields. Perhaps the most prominent cause of result discrepancies is the different ways in which the sites chose to map original data fields into fields that make sense to users trying to query the database. The MESL data dictionary contains 32 fields, far too many to effectively present to any but the most sophisticated user. Each implementation made its own decision on how to group sets of fields into a single user-searchable index, and what to call each of these combined indexes. Because of this, users visiting the different sites will be presented with different indexes over the same underlying content.

    In many cases it is fairly easy to guess which fields were combined into which index. However it is not at all obvious which fields were indexed to form the virtual "keyword" field; there are an enormous number of possible ways of forming this index (from combining prominent fields like subject, description, and title to relying completely on the words within the label field). Careful choice of which fields to index for keywords might help prevent problems like finding an artist named "Child" when looking for portraits of children. Because most of the "simple" or "quick" searches favor the keyword approach, it is likely that most casual users would get different results from the various implementations.

    This issue of how to combine fields for presentation to users can be a fundamental one for museums contemplating an interactive database for visitor access. The MESL experience of trying to make 32 fields understandable to the average user pales in comparison with the museum that must boil down scores of fields from a collection management system into something easily digestible to the user, particularly if the user is unfamiliar with curatorial language.

    Another cause of divergent results had to do with the different ways in which the various search engines work. Most prominent among these is whether all matches must start exactly the same beginning from the left side of any field, whether the system looks for character-strings or whole words, whether they perform stemming, truncations, etc. Searching design decisions such as this can drastically affect search results, as we have seen in the case where a search for oil paintings picked up the word "soil".

    Security

    The security environment restricts access to the database. Although this can take many different forms, we can loosely group it into three categories: proprietary systems, general access controls, or authentication (see Environment Table 4). Unlike other practice environments, the security environment tends to discrete – where the choice of system embodies a set of defined procedures (although some systems can include "features" found in others). Environment Table 4 lists the basic security environments.

    Environment Table 4. Security

    Outline of Cost Centers and Primary Data Source

    Cost Center

    Description

    Source

    1. Security Choice

    Choose both the kind and level of security for the system. There are a range of different possible systems and combinations. Additional layers or components can usually be added (with costs), although the fundamental structure usually limits the range of possibility.

     
  • Proprietary Systems
  • The restriction of direct connections through limited physical access. This can be done through a closed network system, common in many early computer laboratories, or through proprietary network operating systems that only recognize specific protocols.

    UTR, UPR

  • Access Control
  • The limiting of access to the database through a general network domain restriction.

    UTR sections 2.1.5, 2.1.6

  • Authentication
  • Limiting access through restricted physical network access.

    UTR, LS

    Notes: MTR = museum technical reports, UTR = university technical reports, OR = other reports, FGT = focus group transcripts, LS = list server communication, MPR = museum proposals, UPR = university proposals

    Many of the original MESL delivery formulations sought to use "closed" systems (e.g. computer laboratory running proprietary systems). Some early efforts (release 1) reported on these efforts. At most institutions proprietary systems were abandoned for more open web-based systems. Efforts to control access to the database varied (with varying degrees of success). Section 2.1.5 of the technical report summarizes the local security efforts.

    Most institutions reported using an IP domain control, and most did not report "problems" with their security system. This form of security is quick and easy to implement, and only requires a list of valid campus domains or IP addresses. But though IP access control is a quick-and-easy solution, it poses serious problems for a true production-level delivery system. Most IP address control is based upon directory-level security. The design of the systems employing such controls is critical since the directory structure determines access. This means different image sizes and/or text elements need to be stored in different directories based desired level of control rather than upon logical arrangement. For example, a university might want to control access to all images bigger than thumbnails but allow anyone to see textual descriptions. This would mean that thumbnails and text would reside in an uncontrolled-access directory while all other images would be in a controlled-access directory.

    A second issue is that groups of IP addresses are usually over-inclusive of the valid authorized user population in some areas and under-inclusive in others. Commercial entities leasing campus space, private technology-transfer spin-offs, alumni dial-up access, and other groups that might not be valid members of a "campus community" (as defined within a licensing agreement). However, they would often be included within the campus IP domain (and in many cases no record is kept that can be used to isolate these from student, faculty, and staff use). Much valid use comes from outside the core campus domain, in the form of satellite campuses and programs in other cities, students and faculty who dial-up through their own internet service providers, faculty on sabbatical at other campuses, etc. Though a campus might try to create a list incorporating most of these other valid external IP addresses, managing such a fluctuating list would quickly become unwieldy.

    Mid-way through the MESL project, several of the campuses began to plan experiments to implement more sophisticated means of access control. In the second year of MESL, Illinois added login and password access to supplement IP access as a way of serving those outside their core IP cluster. In 1997 both Michigan and Columbia began experimenting with a required authentication login and password for most MESL users. Michigan employed an already-developed database of all campus users and authenticated against that database. However, such implementations require significantly larger commitment of resources.

    It is clear that, in the long run, simple IP access control will be insufficient to support delivery in a way acceptable to most rights-holders. More sophisticated methods need to be found, primarily based upon users rather than upon workstations. Most of these methods will require the universities to keep more careful track of what group each of their users belongs to (so that they can isolate alumni or drop-outs, so that they can grant access to some material solely to those enrolled within a particular course, etc.). And, because of privacy concerns, universities will probably have to maintain authentication systems based upon this level of information about their users, even when distributors are delivering licensed material directly to members of the university community. Universities like UC Berkeley are experimenting with public key encryption and digital certificates to try to solve this problem.

    Usage Environment

    The usage environment addresses how individual images or sets of images are acquired for a particular purpose. It consists of four basic elements: outreach to inform endusers of the database, training in how to use the database, support for ongoing use, and actual access. Environment Table 5 summarizes these areas.

    Environment Table 5. Usage

    Outline of Cost Centers and Primary Data Source

    Cost Center

    Description

    Source

    1. Outreach

    The needed effort to promote and educate potential endusers about the availability and the capabilities of the system.

    UTR sections 2.2.1

    • Interest

    Relevance of the database

    OR: UIUC surveys and Cornell Casual User Survey

    1. Training

    Train endusers on the system. This needs to take several different forms – training for faculty who would want to use the system in the classroom, training for students who would want to use it for research or coursework. Faculty training requires a more extensive effort given what needs to be done in the classroom.

    UTR

  • Support
  • Support includes two sets of distinct activities. The first is training (or additional training and hands-on walk through) on using the system. The second is further technical development of the system and system tools.

    UTR sections 2.2.3

  • Usage
  •    
    • System Access

    The kinds of available access are critical issues for determining how the database is accessed and used.

    OR: UIUC, log files

    • User Interface

    The kind of user interface is a critical issue for determining what the naïve enduser can do with the system (and how the informed enduser can interact with it).

    OR: Cornell

    • Search Tools

    The kinds of tools provided to the enduser for searching and manipulating the database. This includes the number of searchable fields and the kinds of tools.

    UTR, OR: Cornell

    • Retrieval

    What is the form of retrieval.

    UTR sections 2.2.2

    Notes: MTR = museum technical reports, UTR = university technical reports, OR = other reports, FGT = focus group transcripts, LS = list server communication, MPR = museum proposals, UPR = university proposals

    Enduser economic issues are difficult to ferret out. Infrastructure appears to be the biggest concern. Existing electronic classrooms appear to be limited by their ease of use and functionality. Public access via public terminals and the internet offer new challenges to distributors. The costs of providing or improving infrastructure – especially as it applies to digital imaging – is difficult to ascertain.

    The formal costs for providing the database to endusers fall into the first three areas: outreach, training, and support. Outreach, under MESL, mainly centered on staff introducing the MESL distribution concept and technology to faculty and providing special seminars and training sessions. Training required walking individuals through the system. A final critical issue is the costs of support. Section 2.2 asked about outreach efforts to encourage faculty usage, direct support for instruction, and additional effort need to further develop the system for faculty and endusers. While there appears to be a significant investment in outreach and education at all MESL sites, the actual usage outcome (developing courses, etc.) does not appear to be high. A number of factors could be at work, including the lack of key images to the barriers imposed by the lack of, or problematic infrastructure. This needs greater exploration.

    Assessing the cost of formal usage is more difficult. As reported above, many of the applications include department and faculty involvement, but the technical reports do no represent them. Anecdotal information from individual faculty suggests a large time commitment required for course setup using digital images (one reported 200 hours). These large time commitments are, however, probably one-time expenditures (presuming the image database remains relatively stable). Student or casual user access is also difficult to define. For the most part, formal MESL design by the majority of image distributors looked to general web-based architectures as a means to manage the variety of different machine architectures on the university campus. The Cornell "casual user survey" notes that a number of non-students accessed the system looking for specific information. The best way to begin to assess the resource requirements is probably to explore the log file usage patterns. At another level, access or ownership of machines capable of using the images could impact broad usage of MESL images. UIUC data suggest that 40% of students who used MESL as part of their course work did not personally own computers capable of accessing and utilizing digital images, which meant that they had to go to an on-campus facility to work with the images. Even those with the physical capabilities to access the images could be limited by how they are connected to the system. Issues such as bandwidth at off-campus or dial-up sites could slow image access to the point where system functionality was negated.

    Technology Infrastructure

    The technological infrastructure is the ambiguous connections between practice environments or activity clusters that are difficult to assign ownership to. Infrastructure includes issues such as physical space, communication modalities, shared computing environments, information network or highway connections, and time. The technology frames what is feasible. It arranges objects and machinery and shapes the kinds of relationship producers, managers, and users have to them.

    For example one critical area that is impacted by the changing character of the technological infrastructure is the fundamental concepts of exchange. An object’s representation constitutes a value that is at once distinct from the actual object but also strongly correlated to it. The physical representation helps to determine the core units in the technology infrastructure. There are several distinct representation layers. At minimum there are the "print" values such as transparencies that can be associated with print publications, there are also other analog images such slides, and there are digital images. Values as identified by the individual institution that "own" the rights to the object or want the rights could be attached to these layers. For example, each provider institution has a price point for the sale of a slide or rental of a transparency (and, conversely, each enduser has price point for the purchase or rental). The values of digital representations are only now being explored. This most apparent issue being the notion of "image quality." In MESL, the definitions of "quality" was clearly different for museums (using an analog framing) and universities (using a digital framing). In practice, the distinction between a high-resolution digital image and a low-resolution video display quality image was made. The substantive issue was framed by the ability to easily manipulate digital images to produce "derivatives" with varying characteristics (e.g. thumbnail sized to full screen representation). A significant secondary issue is the subtle technical issue of what is represented – is it the object itself, or is it a surrogate for the actual object such as an analog slide (Trant, 1996c). Any discussion of digital distribution needs to disentangle the varying values found in operational concepts.

    Determining the formal infrastructure requirements is critical in any evaluation of the economic viability of a project because it provides a baseline measure of the minimal formal context needed for the practical implementation and the successful achievement of a project’s goals. However, actual infrastructure costs are always difficult to determine. For example, digital distribution presupposes that a set of reasonable technological capabilities already exists and is in place: there are objects; there are mechanisms to digitize them in the appropriate form, there is a means to transmit them, and mechanisms to retrieve them. While transmission under this model ostensibly means using a seamless "electronic digital network," transmission can also mean using the "sneaker network" where digital records are handed from entity to entity. Under MESL, the appropriate transmission of both the set digital objects from the museums and the raw data to universities was tape (or CD-ROM) and the mail, and the appropriate transmission of individual digital entities was the electronic network. This means that while digital networks are required for the delivery of specific end-products (an individual digital entity) they aren’t necessary for all parts of the process. Furthermore, an end-product delivery can be achieved through direct dial-up access

    The formal network thus needs to be evaluated in terms of its functional relationship to those that use it. For example, MESL operated under a dual set of presumptions about institutional technological capabilities. MESL assumed that the Universities had a technological infrastructure capable of delivering digital images. In their application to join MESL, each institution needed to outline their technical capabilities and competencies. On the other hand, not all museums were expected to have such infrastructure. Among provider participants, digital-imaging experience ranged from fully implemented digital projects to a museum whose personnel did not have electronic mail. This created different experiences in actual procedures such as digitization and the transfer of images. Institutions with fully developed digital projects had specific procedures to follow and established cost structures, while less experienced institutions had to climb a learning curve on the digitization process. These differences also resulted in varying calls for formal digital standards, where less experienced institutions wanted procedures and image quality specifications spelled out. The different experience also translated into varying resource expenditure on MESL as whole. At universities, the infrastructure issues in image delivery include changes in primary network protocols that impacted the character of the physical network – sites that had proposed Novell solutions shifted to more open TCP/IP and "web" applications. These evolutions in formal organization reflect, in part, general changes in the technology. At the time MESL was being implemented, a new commercial interest in the "web" and whole sets of "web-based" applications were being deployed.

    Finally, the technical infrastructure needs to address the general costs of actual physical capabilities. As reported above, all universities reported having digital classroom, but all saw limitations in these classrooms affecting the overall adoption of the technology. Physical electronic classrooms clearly had different capabilities ranging from interactive networked terminals to simple single network connection to a general display. The different capabilities clearly shape the possible classroom access to MESL and influence instructor’s willingness to teach with MESL material.

    The technological infrastructure establishes formal limits to what was possible for MESL. Many participants identified the kinds of infrastructure questions the project raised. This ranged from looking for new management systems at museums, to rethinking digital classrooms in the universities. On the museum side MESL suggests that general distribution of digital images could be viewed as a collateral benefit rather than a primary one. Digital images – at the level of quality used in MESL – offered internal benefits for collections and management systems. The distribution of these "screen quality" images also led students close to MESL sites (Washington DC area) to travel to the museum to actually see the artifact of interest. For universities, the definitions of access, enabling faster access, and limiting access were critical issues that appear to be progressively explored.

    The range of infrastructure costs and benefits are difficult to ascertain, but need to be acknowledged. For the purposes of the Mellon Project, we will simply identify the central infrastructure issues that are critical for digital entity distribution. This includes real objects (such as physical computers and digitization devices, digital networks and connections, and actual classroom) as well as critical concepts that affect distribution (such as what gets represented). We will take as natural constants (and therefore ignore) issues such as power consumption or physical building space.

    Social Organization

    The institutional organization clearly impacts formal implementation and shapes the character of outcome. Under MESL, there was a clear effort to select different kinds of institutions – public and private organizations – as participants. All sites stressed varying degrees of participation from multiple internal units. What specific departmental had overall MESL responsibility and the ability to mobilize resources appear to be critical factors for understanding the local outcome of the project.

    Four private and three public institutions were selected as image providers. Each had a distinctive character to their collection, and brought different technological sophistication to the digitization process and to the creation of fixed textual descriptions. Proposals from participant institutions to MESL originated from various internal departments – information services, education centers, new media initiatives, and the registrar. In the museum proposals, a number of additional departments were identified as participating units. Importantly as the project progressed, operational control was shifted in three institutions to another internal unit. Table 5 summarizes these data for Museums.

    Table 5. MUSEUM – Proposal and Operations: Lead and Participant Units

    LOCATON

    Submitted Proposal

    Actual Operation

    Change notes

     

    Sponsor

    Other Units

    Lead Unit

    Other Units

     

    Administration

     

    5

     

    3

     

    Collections

     

    1

    0.5

    2

    NGA – 50% publication

    Computer/IS

    4

    1

    3

    3

    GEH, FMCH, LOC

    Curatorial

     

    2

     

    2

     

    Education

    1

    4

    1

    2

    NMAA – new media

    Legal

         

    3

     

    Library

     

    1

    1

     

    MFAH – education

    Photography

     

    3

     

    3

     

    Publications

    1

    1

    0.5

    2

    NGA – 50% collections

    Registrar

    1

    4

    1

    2

    HUM

    Research

     

    1

     

    1

     

    Notes Data sources come from named individuals or units as part of the original project proposals and from the technical reports (section 1.4 staffing).

    Named participants from other units appear to cover the major operational entities with varying participation among institutions (range of 2 to 7 participating units). The most interesting observation is that legal counsel was not identified as a required resource in any of the proposals. In practice, legal counsel became a formal part of the MESL operations at three sites – the other sites self-consciously excluded legal counsel by only distributing images they knew they had full rights to. Thus, while all MESL producers did not formally involve legal counsel, it was clearly part of the production process.

    Among the image distributors, there were three private and four public universities. Among these institutions, there were distinctive differences in size and each had a distinct technological sophistication in digital imaging. Because the primary mission of MESL was the development of "practical mechanisms to distribute electronic images," MESL required universities to demonstrate that digital images would be used in the classroom as part of their application. Six project proposals originated from the library. The seventh came from a department. Table 6 summarizes these data.

    Table 6. UNIVERSITIES – Proposal and Operations: Lead and Participant Units

    LOCATON

    Submitted Proposal

    Actual Operation

    Change Notes

     

    Lead Unit

    Other Units

    Lead Unit

    Other Units

     

    Department

    1

    6

    1

    0

    UMD

    Information Technology

    0

    7

    3

    7

    Columbia, UMI, UVA

    Library

    6

    7

    3

    7

    American, Cornell, UIUC

    Notes. Data come from named individuals or units as part of the original project proposals and from the technical reports (section 1.4 staffing). Individual Faculty members were named when a list of courses was provided.

    In the course of the project, control of three library projects shifted to computer and information services units. Significantly, while all proposals invoked departmental sources, practical operations did not need their direct participation. In fact, departmental faculty appears to be construed as client endusers for the delivery team.

    A second issue is in the area of personnel commitment (see Table 7). The issue of personnel can be used a crude indicator of overall resource commitment by the local institution. Places with a low body-count could be construed as having little institutional support, while those with larger staffing suggest greater support.

    Table 7. MESL Personnel by Institution

     

    YEAR 1

    YEAR 2

     

    personnel

    personnel

    Personnel

    Personnel

     

    over 50%+ FTE

    Under 50% FTE

    Over 50%+ FTE

    Under 50% FTE

    MUSUEMS

           

    Fowler Museum of Cultural History

     

    8

     

    7

    George Eastman House

     

    2

     

    3

    Harvard University Art Museums

     

    6

     

    6

    Museum of Fine Arts, Houston

     

    13

     

    11

    Library of Congress

     

    6

     

    5

    National Gallery of Art

     

    6

     

    5

    National Museum of American Art

     

    8

     

    7

    UNIVERSITIES

           

    American University

    1

    9

     

    10

    Columbia University

     

    11

     

    11

    Cornell University

    1

    6

     

    4

    University of Illinois, Urbana-Champaign

    2*

    4

    2*

    4

    University of Maryland

    3*

    11

    3*

    11

    University of Michigan

    2

    8

     

    12

    University of Virginia

     

    7

     

    3

    Notes. * - Students.

    All sites include those who simply participated in "advisory roles." These distinctions would decrease the number of actual working participants.

    Source: MESL Technical Reports. Includes both those reported as MESL and those "Doing the work of MESL."

    The reported numbers appear to formal under-reports of how many resources were actually committed to the project. The cost of doing the work of MESL appeared to decline during the second year at all sites. An important understanding of overall personnel commitment for MESL is the fact that much of the project was completed with a low number of individuals employed over 50%. Many of the individuals identified as less than 50% had only a few hours for the entire year committed to the project (different year 1 and year 2 commitments).

    The distinction in participant unit layers and personnel resource commitment gives us a frame for deciphering differences (if any) in MESL cost structures. The variety of institutions involved and the general success of the project suggests that digital image distribution is feasible for relatively low costs. However, one of the critical questions that needs to be explored is the differences between low resource commitment and overall quality of either image provided or interface. It permits an assessment of the practical differences and outcome of MESL. For example, viewed institutionally, the shifts in unit control could represent a cosmetic change in responsibility or de facto project control, or the change could be suggestive of a larger shift in organizational interest. Institutional sponsorship, defined by number of active participants could also be significant for assessing how MESL was played out.

    SUMMARY

    The economic examination of MESL is still in process. The next step will be to attach summaries of resources needed and other costs associated with each cost center. The analysis is difficult not only because of MESL’s innovative nature but also because of its formal design. The many-to-many experiment with varying institutional characteristics and different data collection points makes simple mechanical comparisons problematic at best. To date, many of the critical technical hurdles have been addressed including defining methods to manage problems with data quality and comparability.

    A comparative examination of analog image distribution/delivery systems (slide libraries) is in its opening stages. The analog model is an important comparative case because historically slide libraries have been the primary method for the mass distribution of cultural heritage information to the educational community. We have grouped cost centers in ways to expedite comparisons between analog and digital versions of each. We expect to find that a number of the cost centers for a slide library closely duplicate efforts already undertaken by museums, and that these costs might be greatly diminished in a more cooperative environment. For example, the slide library’s cost center for textual description of the object depicted in a slide might be replaced by a payment to a museum for conversion of their own textual description of that object into a form acceptable to the slide library. Appendix 3 – THE ANALOG DISTRIBUTION CHAIN –contains more details.

    To date the evidence suggests that the digital distribution model should not be viewed as a direct replacement of analog distribution systems. The lack of substantive infrastructure in the academic environment (e.g. digital classrooms, appropriate workstations and labs) and useful tools to help manipulate image sets makes the functional use of these system problematic. In this study we will identify these barriers to widespread adoption of digital distribution/delivery systems, as well as advantages that have enticed faculty and students to use these systems despite their drawbacks. But for the foreseeable future, the digital distribution model should be understood as a supplement, rather than a replacement for analog distribution systems for cultural heritage information.

    We also expect that a study of value assigned to digital images by users will be relatively low. We surmise that museums should not expect an income stream from sales of images in this format, and (from a monetary standpoint) should at best hope for cost-recovery. But we also expect that digital distribution will yield enticing benefits not related to a potential income stream. This can take the forms as diverse as: increasing the museum’s public profile and attracting new use and attention; in promoting new capabilities, work procedures, and staff relationships within the museum itself; encouraging standard practices and cooperative ventures between museums.

    The Mellon Project will be completed in early 1998. 

    REFERENCES

    Albrecht, Kathe. 1995. Digital Imaging on Campus: An early look at the Museum Educational Site Licensing Project, SECAC 1995 Abstracts, Visual Resources, vol. 22, no. 4, p. 12.

    Bearman, David. 1996. New Economic Models for Administering Cultural Intellectual Property, A paper presented at the Digital Knowledge Conference, Toronto, Ontario, February 7 and EVA Florence, Italy, February 9.

    Besser, H. and C. Stephenson. 1996. The Museum Educational Site Licensing Project: Technical Issues in the Distribution of Museum Images and Textual Data to Universities, a paper presented at the Electronic Imaging and the Visual Arts Conference (EVA '96), London.

    Giral, Angela and Jeannette Dixon. 1995. The Virtual Museum Comes to Campus: Two Perspectives on the Museum Educational Site Licensing Project of the Getty Art History Information Program. A paper presented at the Conference of the International Federation of Library Associations and Institutions (IFLA), August.

    Lebowitz, L. 1996. MESL Brings Museum Images to Universities, cover article for October, issue of the University of Michigan's "Information Technology Digest"

    Trant, J. 1996a. Enabling Educational Use of Museum Digital Materials: The Museum Educational Site Licensing (MESL) Project, A paper for the Electronic Imaging and the Visual Arts Conference, Florence, Italy, February 8-9.

    Trant, J. 1996b. Exploring New Models for Administering Intellectual Property: The Museum Educational Site Licensing (MESL) Project, A paper prepared for the 33rd annual Data Processing Clinic, University of Illinois at Urbana-Champagne, March 25.

    Trant, J. 1996c. New Models for Distributing Digital Content, A paper for the Data Processing Clinic, University of Illinois, March 25.

    Trant, J. 1995. The Getty AHIP Imaging Initiative: A Status Report, Electronic Imaging and the Visual Arts (EVA), The National Gallery, London, July, 1995. Also appearing in Archives and Museums Informatics, Cultural Heritage Information Quarterly, Vol. 9, no. 3, 262-278.

    Trant, J. 1995-96. The Museum Educational Site Licensing (MESL) Project: An Update," Spectra, Winter.

    Trant, J. 1994-95. The Museum Educational Site Licensing Project, Spectra, Winter, pp. 19-21.

    DOCUMENTS

    MESL: Goals and Objectives, Getty Information Institute, 2/22/95. [http://www.ahip.getty.edu/mesl/about/docs/fact.sheet.html] copy date, 12/16/96.

    MESL: fact sheet, Getty Information Institute, 10/18/95. [http://www.ahip.getty.edu/mesl/about/goals.html] copy date, 12/16/96.

    MESL: Call for Participation. Issued September, 1994. [http://www.ahip.getty.edu/mesl/about/docs/call.short.html}

    MESL: Call for Participation: Museums and Galleries, Issued September 1994. [http://www.ahip.getty.edu/mesl/about/docs/call.museums.html]

    MESL: Call for Participation: Higher Educational Institutions, Issued September 1994. [http://www.ahip.getty.edu/mesl/about/docs/call.education.html]

    MESL Proposals from each institution.

    MESL Meeting reports.

    Mellon Project Proposal: Museum Educational Site Licensing Project: Studying the Economics of Network Access to Visual Information, 11/96. [http://www.sims.berkeley.edu/~howard/mellon/]

    APPENDICES

    Appendix 1 – METHODS

    Appendix 2 – BACKGROUND

    Appendix 3 – THE ANALOG DISTRIBUTION CHAIN

    Appendix 1 – METHODS

    One of the unique features of the MESL experiment was the willingness to create a situation where standards for the digital distribution of images could emerge – rather than imposing them and seeing how they worked. By design, MESL was a many-to-many experiment – multiple museums sending a large set of digital images to multiple universities for distribution. More importantly, there was an effort to introduce a heterogeneous mix of institutions with different characteristics, different experiences with digital imaging, and different visions of image delivery. This willingness to encourage and explore the possibility of emerging standards within a broad mix of institutions makes it difficult to statically assess costs estimates and outcomes.

    The technical problem of the economic evaluation is to reconstruct the full range of activity – both formal and informal – needed to accomplish MESL. However, from the beginning, the economic evaluation was confronted by practical problems in data collection. First, although the digital objects were similar, their production and distribution environments at each institution were theoretically different. This meant that specific commonalties and cost centers along the production and distribution chain (from the museums' creation of the images and accompanying text, to their distribution to the universities, to their to their use by individuals) needed to be derived from diverse project implementation designs. This heterogeneity makes standardized data gathering problematic. Second, at all institutions, MESL was understood an experiment in the electronic distribution of digital images and data. MESL participants were thus on a steep learning curve, caught in a web of solving specific technical issues in order to accomplish the overall goals of the distribution. Thus, although individual sites were asked to keep accurate logs of what was happening during the course of MESL implementation, most were not kept. Understanding much of the MESL experience therefore has to be recreated from memory. Finally, MESL had a significant learning curve on how to effectively distribute digital images. Thus many of the costs are likely to be higher than they would be during a true production mode.

    Recognizing this institutional heterogeneity and experimental mindset, the MESL-Mellon evaluation team identified data collection points experiences could be triangulated. Figure 3 outlines the data collection relationship between the economic evaluation, MESL, and primary data sources.

    Figure 3. Data Sources

    For the economic evaluation, the primary data collection device was the technical report. It was jointly developed by MESL management and the Mellon Project and solicited from each participating institution as part of their project reporting. Supporting data came from the MESL project archives, site visits, focus group interviews, surveys of MESL participants and endusers, and web server log files.

    1. Technical Reports. The reports review the implementation requirements of MESL and include the documentation of the associated economic costs. Parallel reports were developed for museums and universities. Each has three sections: an institutional profile, technical implementation, and reflections on experience. The profile documents the resources of the institution, general procedures, and staffing of MESL. The final section asks open-ended questions so project team members could expand on their experience.
    2. The heart of the report is the technical section that documents procedural moments. For museums, this includes information about collection management systems, content selection, and image and text processing. For universities, this includes system architecture, data preparation, functionality and support.

      Each question of the MESL technical report can be linked to specific production environment and mapped to a logical procedural moment in the digital imaging distribution path. For example, sections 2.1.4 in the Museums report asks about the digital imaging process. The series of questions under this section include interrogation about past experience, prior resources, and MESL experience and resource commitment. On the other side, Section 2.1.3 asked the universities about the data preparation and loading requirements of MESL onto their systems. Questions under this section interrogate resource needs and expenditures for different kinds of data preparation (e.g. images and both structured and unstructured text). The cost reports include an assessment that ranges from "highly accurate" to anecdotal estimates. The individual moments can be collected into functional "cost centers" in the workflow.

    3. Cost Report from the central processing facility. The University of Michigan’s central processing facility submitted their own cost report on their activities in text and image aggregation and correction. It evaluated the types of problems and included costs estimates. This report provides insight into the technical hurdles confronted by the museum image production environment. It provides the core data for understanding the Entity-Processing Environment.
    4. MESL archival data. These data include project announcements, request for applications, project proposals, MESL electronic lists, and published reports.
    5. Focus Group Interviews. Group interviews are useful devices for eliciting data on individual experience that otherwise might go unreported. The group setting allows "personal" experience to be expressed in a situation where they might be collectively understood as common ones. The focus groups were done during the final MESL project meeting.
    6. End User Surveys. Objective measures for the enduser environment come from a cooperative effort of several teams working with different instruments that target distinct features of MESL. The University of Illinois at Urbana-Champaign is charged with an evaluation of the classroom experience at the MESL academic sites. Cornell University (CU) is conducting an evaluation of enduser response at the MESL web sites and a survey of MESL participant experiences.
    7. Site Visits. The Mellon team members have conducted six formal and informal site-visits to MESL institutions to evaluate the impact of the project. These data are informative because they allow participants to discuss and demonstrate their own MESL experience. Specific meetings with faculty during these visits to universities provide important clues about the creation of courses and impediments to use.
    8. Server logs. A few of the MESL university sites that use Internet web servers have made their log files for a single academic semester (Spring 1997) available for further investigation. Log files provide useful information about how the image server was accessed. Data mining of these files can reveal when the site was accessed, where users came from, where they entered, and where they left. The files may provide some insight about what functions were tried, in what order, and how long they took.

    The evaluation recognizes the variation in kinds and qualities of the data. Interpretation requires linking "hard" or "objective" data to softer, experiential or qualitative evidence. The sections in the technical report can thus be mapped to supporting documentation. Site visits and focus group interviews help to interpret the technical reports. Supporting evidence for data in the background sections of technical reports are derived from the request for applications and proposals submitted by each applicant. Supporting material for the technical implementation come from meeting notes and list-servers, as well as data from the central processing report. The reflection section of the technical reports receives support from the various participant surveys. These relationships are illustrated by Figure 4.

    Figure 4. Data Sources Object Relationships

    The formal design of MESL saw the deliberate bringing together of different institutions with distinct strengths and interests. Although general cost center analyses and cost trajectories are instructive, the disparities between institutions are as important as their similarities. The heterogeneity of institution means that any individual site can be viewed as an archetypal example of a specific type rather than a member of a collective. The analysis thus has to be both an intra- and an inter-institutional exercise situated within context and examined in terms of local social organization and local culture. This requires a two-step examination of costs that not only compares the data quality across institutions (correspondence) but also its link to accounts in individual experience (coherence).

    Appendix 2 – BACKGROUND

    The Museum Educational Site Licensing Project (MESL) wanted to "define the terms and conditions under which museum images and information can be distributed over campus networks for educational use" (MESL: Goals and Objectives, 2/22/95:1). The project recognized that networked resources offered to remove "some of the physical barriers to the enjoyment of cultural heritage collections, making them available to wide audiences, including those who normally might never enter a museum building." More importantly, these technologies could change "the nature of teaching and research" because digitized "images of works of art and artifacts – accompanied by textual descriptions – can be used in new, exciting ways, [by] placing works in context rather than isolating them." However, in order for this to happen, "a critical mass of digital information must be available in standard forms" (MESL: fact sheet; 10/18/95).

    MESL was designed as a prototype demonstration of using "digital imaging and network technologies" to "make cultural heritage information more broadly available." It consisted of three objectives:

      1. To develop, test and evaluate procedures and mechanisms for the collection and dissemination of museum images and information.
      2. Propose a framework for a broadly-based system for the distribution of museum images and information on an on-going basis to the academic community.
      3. Document and communicate experience and discoveries of the project.

    (MESL: Goals and Objectives, 2/22/95:1-2)

    These objectives addressed the two fundamental questions about the feasibility and long-term access to digital images. MESL provides us with a unique opportunity to examine costs and uses of digital images delivered over campus networks. The goal of the economic evaluation is to identify and evaluate the technical infrastructure requirements and the social and economic resource needed to accomplish the MESL project. Understanding the costs associated with various steps in the MESL distribution process is critical, not only for understanding the relative success of MESL, but also for framing the future direction of the digital image distribution concept.

    An analysis of costs requires a model of the digital distribution of images. As a higher order representation, it should establish a cognitive map, or schema, that provides a theoretical picture of what should happen. The model needs to differentiate between the inter-organizational entities, intra-organizational layers, and the distinct social units within them. It should locate the technological infrastructure and outline the central technical processes and activities in the social organizational regions. Finally, it has to identify actual work trajectories and determine what actions are contingent on other sets of sub-processes. Such a map permits a comparison between the original project design, what generally took place, and what transpired at any specific institution.

    The first MESL objective defined a feasibility study – identifying whether the networked distribution of a large number of images from museums to universities was possible. The formulation offered a basic distribution model where digital images and their accompanying text documentation were created by museums and distributed to universities who made them available to their users for use in the classroom and for research. As an exercise in feasibility MESL has two basic procedural elements: image and text producers (represented by museums), and the distribution of these image and text by universities. The packaged distribution of digital images of museum objects and their associated text required the creation of a new digital entity, a package that links distinct data elements. In MESL this consisted of a digitized image and text data but digital entities could also include audio and/or movies. The technical hurdle in MESL was combining two sets of digital information – relatively new digital images and text data held primarily as legacy data in local information systems. Extracting legacy data from systems designed around a local culture into a shared database imposes a certain degree of standardization.  At minimum, this requires determining the appropriate classification categories and translating the local terminology into a shared vocabulary. In MESL the shared vocabulary needed to be negotiated between seven distinct providers and the Getty AHIP vocabulary for textual information was proposed as the shared model in the initial project proposal.

    Figure 5. MESL Delivery Model

    The second objective sought to define the underlying character of the relationship – the conditions for using the images. This reflected the practical concerns of a production model, where digital images could be efficiently distributed. Production is predicated by two sets of criteria. First the permission to use and access the images. Second, interest in the provided images. The first criteria establishes the formal "terms and conditions" of usage, the second locates the individual preferences that make them relevant. The terms and conditions are the primary conditional framing of MESL that gives museums permission to develop digital images and access to them by universities. Underlying the delivery system is the actual usage of any image as determined by the selection of the content provided by museums and the particular interests that make the relevant. Figure 5 illustrates the relationships in the basic MESL delivery model.

    However, this museum-university framework has natural, practical limitations. At the very least, the delivery of images requires a third element: the enduser. This creates a theoretical delivery model where producers simply deliver digital images to deployment site who then pass them on to endusers (see Figure 6). Under this model there is a direct one-to-one relationship between the terms and conditions, delivery model, and preferences.

    Figure 6. MESL Theoretical Delivery Model

    In the practical world, this third unit alters the basic terms and conditions of image use and distribution. While the fundamental principal in the terms and conditions of use presume that it is possible to control enduser usage, actual university practices limit the kinds of formal control over the enduser. This creates a natural asymmetry between the formal terms of usage and actual practice. The management of this asymmetry was sought in the underlying technology of the digital distribution where control over the physical access to the image was possible. The three elements – demand, application, and conditions formed the basic outline of the overall distribution concept (see Figure 7).

    Figure 7. MESL Practical Delivery Model

    One of the unique features of the MESL experiment was the willingness to create a situation where standards for the digital distribution of images could emerge – rather than imposing them. This willingness to explore the possibility of emerging standards was immediately impacted by the many-to-many relationship. The high number of digital objects, combined with a number of sender and receiver institutions with different digital imaging skill levels increased the possibility for file errors in basic file structure. While each recipient could theoretically manage these differences (and errors), there was an early recognition that the effort to create a minimally usable data set would be duplicated. While several solutions were proposed, the MESL project opted to have a single center conduct basic file maintenance. The responsibilities included: basic quality assurance file checking of digital images and associated text data sets for noticeable errors (i.e. delimiters), standardization of the data sets into a uniform digital appearance, and to package and deliver these data to the universities. This model recognized that the integration of data elements into functional digital entities and the aggregation of the multiple data sets would require each university to perform additional processing in order to develop their local database. Each university then used the corrected data to implement their own database and delivery systems.

    These objectives and modifications established MESL’s distribution system and its operational frame (illustrated in Figure 8). There are physically distinct institutions: museums that produce the images and text, a central processing facility for error checking and quality control, universities that deliver the images and text, and endusers who access the images for particular urposes (illustrated by items 1a-1e). Activity within each zone is framed by (2) the terms and conditions of use, and (3) the usage preference and relevance.

    Figure 8. MESL Actual Distribution Model

    The economic evaluation of MESL has to recognize the distinction between the physical elements of digital image distribution and their social location. The model used by MESL constitutes one of many possible distribution methods. Regardless of their social location, the delivery of digital objects requires three fundamental units in (1) the physical delivery path (a. image creation, b. distribution, and c. use). Structurally there are two intervening elements that need to be formally worked with: (1d) image checking, and (1e) controlled access by end users. A critical assessment of costs associated with this path needs to address the two framing issues of digital image usage: (2) the formal terms and conditions of use, and (3) the usage preference and relevance. The formal terms and conditions frame locates (2a) museum permissions to digitize images and (2b) access to these images at universities. At some level, these elements involve direct costs for digital image, at another point they are simply a concern frames technical implementation. For example, the legal document constrains (2c) the types of usage an individual has and impacts the physical design of the control system (1e). Usage preferences and relevance directs (3a) the selection of images, (3b) the kinds of courses than use the images and (3c) what image are actually retrieved. For the most part, usage preference and relevance are not formal costs, but they clearly direct the amount of time and energy spent developing access to these images such as courses.

    This basic schema allows for the exploration of the practice environments along the digital image delivery pathway.

    Each practice environment offers an encapsulated institutional component of the digital image creation and delivery process. Collectively these environments form a distribution pathway. This basic schema allows for the characterization of the functional digital image delivery pathway (Figure 9).

    Figure 9. Basic Digital Image Distribution Model

    While the structural order mirrors that of MESL, the functional component elements for the economic analyses are different in several respects. Under MESL each environment was institutionally discrete – museums produced, universities deployed. In other models responsibility for processes can reside internally to an institution or they can be shared or contracted out among several institutional sites. However, as in MESL, the terms and conditions of use, and the preference and sense of relevance of endusers will frame activity within each zone. Organizationally, the functional model first identifies a processing environment. This environment includes both MESL’s central processing facility and a portion of the work done by universities – the integration data elements into functional digital entities, and the aggregation of the multiple data sets into a cohesive sequential database container. Second the functional model adds a formal security environment because the implementation of any security apparatus tends to engage actions that are distinct from deployment or use. More importantly, during the later stages of the MESL, several institutions began to move away from simple access controls to more formal security implementations (e.g. authentication). These change the costs of the infrastructure requirements.

    Each practice environment encapsulates a production process that includes multiple component tasks. The task of the economic evaluation is to locate all these elements and determine their relationship and relevance. Identification requires evaluating their economic impact since the cost of these tasks range from continuous direct charges to one-time charges, indirect assessments, or simply technical concerns.

    Appendix 3 – THE ANALOG DISTRIBUTION CHAIN

    A comparative examination of analog image distribution/delivery systems (slide libraries) is in its opening stages. The analog model is an important comparative case because historically slide libraries have been the primary method for the mass distribution of cultural heritage information to the educational community. We have grouped cost centers in ways to expedite comparisons between analog and digital versions of each.

    Every slide library appears to have a production environment where analog images are either produced or acquired. This usually includes some well-equipped, physical space for the capture of analog images. There is also a processing environment where the finished images and their data are checked and entered into a physical record (and any additional information also captured and entered). The deployment environment is the actual physical library and storage space for the analog images. This requires a number of additional resources including check-in/out procedures and "restacking" procedures. There is usually a security environment that restricts access by endusers usually determined by specific status (e.g. faculty, graduate student, or advanced undergraduate working on a project). There is usually a check-in process. There is also an enduser environment where individuals can physically access or use the images. This can range from physical removal of the images from the site, to simply handling individual images on a light tables, or simply being able to view images in a back-lit window. These processes permit a direct comparison to the digital distribution system outlined above.

    We plan to use a number of different data collection devices for our evaluation of slide libraries. Our primary device will be a survey that mirrors the technical reports. Most slide libraries do the work of the digital image producers (museums) and of digital image providers (universities). The survey will include the primary elements from both. We will also collect standard documentation of library procedure, estimates of both the number and cost of producing new images and cataloging them, as well as log data on actual usage. Importantly, many slide libraries are slowly moving to automated records management system – which can include digital versions of the stored image. These additional practices will also be evaluated. To date we have four comparative sites, including two slide libraries at institutions that actively participated in MESL.

    We expect to find that a number of the cost centers for a slide library closely duplicate efforts already undertaken by museums, and that these costs might be greatly diminished in a more cooperative environment. For example, the slide library’s cost center for textual description of the object depicted in a slide might be replaced by a payment to a museum for conversion of their own textual description of that object into a form acceptable to the slide library. This study will provide significant comparative evidence for the overall effectiveness, strengths and weaknesses, in the digital image distribution model.

    NOTES