Intellectual Property and Economics

chairs: Christos Nikolaou and Michael Wellman

Meeting Report

August 27-29, 1997

Pisa, Italy

Introductory Remarks

First Round Presentations (What have you been doing lately?)

Second Round Presentations (What are the pressing research issues?)

Presentations from the Chairs (Answering both of the questions above)

Closing Day's Discussion





Introductory Remarks

  • Purpose for this meetings: Generate topics for further research in Intellectual Property and Economics with respect to Digital Libraries.
  • Output from this group (after the second meeting) - one or more of the following:

  • define common research projects; proposals for joint/complementary funding
  • documents
  • further discussions at conferences

  • Options are open-ended. There is a possibility of a third, EU-supported meeting.

Agenda

  1. Introductions - exploring the diversity of goals, backgrounds, and agendas.
  2. Charter Development - developing research priorities; "What sort of projects should be funded?"; each member should address the topic/question. During these discussions, individual differences will either be resolved or become open questions .

Back to Top

First Day's Presentations

Ann Okerson

Ann discussed several issues, related to the delivery and pricing of information to end users, mostly from the perspective of a large academic library.

Licensing - The LibLicence project <http://www.library.yale.edu/~llicense/index.shtml> identifies language for publisher/library licenses, hoping to provide a standard for other librari es and publishers to use. In today's environment, librarians and publishers spend many hours negotiating the same points, redefining terms over and over. Each license is somewhat different, and enforcing the terms of these contracts is nearly impossible. If the terms paralyze users, (or if users are not legally allowed to be informed of the contract's terms) they will violate them often.

Pricing of Electronic Resources - The end user generally does not pay for the use of electronic resources; where, then, does the funding come from? Ann feels that the producers are looking to the libraries to make up for revenue lost from a drop in indivi dual subscriptions. The libraries, at this point, do not have access to individual funds. Most libraries are keeping the print version of journals, and so are paying much, much more to obtain the electronic version as well. The funding situation for acade mic libraries, of course, is not wonderful to begin with.

Coherent Collection of Print and Electronic Resources - Want the various databases to interact with the rest of the collection. Where is the gateway to the coherence? OCLC, Dialog, overarching software, library catalog..? The library spends a large portio n of professional time instructing users. The maintenance of the digital resources also requires time and money.

Back to Top

Costis Dallas

Costis is working on a couple of relevant projects.

  1. Development of Information Systems for museums. Drawing on research publications, museum catalogs, archives, etc. Developing a semantic model of museum information, looking for interoperability. Need to develop a model of the use of information. Inter ested in bundling entire intellectual element - paper and other necessary references.
  2. Research in Archaeology - Internet Archaeology. There are new services and methods of querying now technologically possible for researchers. How do we make resources effectively available.

So far, legal and economic issues have not been addressed in these projects.

Back to Top

Bernard Rous

Background - Bernard works for ACM, a membership organization whose goal is non-profit scientific and educational dissemination of information about science and technology. They accomplish this through publications and sponsorship of about 50 conferences. ACM is trying to transform to the electronic world, creating digital libraries of its own collections, and trying to reinvent itself as a digital community.

Experience -

  1. Technical Layer Projects. There are two systems being pursued for digital document creation. One creates "digital pages" and the other is an application of SGML, more document-oriented. The two systems differ in many ways. For example, the e xisting technology, workflow and personnel leverage the digital pages approach, which of course has economic implications.
  2. Intellectual Property Rights. What is needed is a copyright policy which makes sense to networked users. Technical safeguards are not worth a large investment, and large-scale abuses will be visible.
  3. Economic Issues. "How to maintain a financially viable publishing mechanism while transforming the mode of production and distribution?" Revenue is at risk from the loss of some individual subscriptions and multiple subscriptions. The multip le-use factor in the electronic world is different, and less certain, than in the traditional print world. As a result, prices for the electronic journals are higher. The support of electronic resources and services are ongoing costs. The savings in distr ibution costs affect mostly the marginal costs, rather than the first-copy costs.
  4. Concerns about Scholarly Behavior. There is a mismatch between author and reader communities of some scientific and technological publications. Some publications are written for, but rarely read. As a result of specialization, there are more publicati ons, with smaller readership/publication. This creates a need for a scientific magazine, which is very costly to produce.

Back to Top

P. Bernt. Hugenholtz

Bernt's institute investigates information law as such; any information - any media (technology-neutral). The issues covered include property rights and legislation, public information law (information access), private information law (press law), informa tion infrastructure, copyright law, and commissioned research.

The Imprimatur project <http://www.imprimatur.alcs.co.uk/frmain.htm> is developing operational copyright management systems. They study relevant legal issues, such as the liability of electro nic intermediaries, user freedoms overridden by contracts, encryption.

As a legal practitioner, Bernt is involved in an intellectual property case, involving the electronic rights of journalists to republish works in electronic form. In Europe, the journalists will probably win, since they have more rights, and the copyright law is author-focussed. The moral rights part of copyright law is stronger and broader in Europe. In the U.S., moral rights are thinner and more waivable.

Back to Top

Pam Samuelson

Pam works on issues of the Information Society, particularly the role of copyright law. There is tension between the desire to maximize information commodities and social order/objectives, such as authors' rights, education, etc. The commodity notion of i nformation is becoming popular in digital libraries; the prevalent technologies are strong on privacy issues and less involved in social good issues. We need to take a holistic approach when creating information policy; access to information is a very gen eral problem, affecting all types of organizations.

At the present time, there are many layers of protection of a work, such as copyright law, contracts, etc. So many issues are unsettled, and the allocation of rights is not all that certain. The layers of protection aren't certain either, and take on an a ura of allocation of rights. In this way, rights are sometimes obtained through the back door. What we need is to look at the big picture, identify the core values, and then develop a core set of rules to support those values.

Pam is teaching a new course <http://www.sims.berkeley.edu/courses/is296a-2/f97/>, in reaction to the Clinton White Paper.

Back to Top

Sebastien Steinmetz

Sebastien has been working on the development of Internet technology in France. In the early eighties, on the technological edge, France developed Minitel, a centralized, single network. Many services were developed for this environment, which has differe nt pricing methods than the present-day Internet. The Minitel services are now being transferred to the Internet, but the transfer is not always easy. Since it is not possible to achieve the same pricing and market structure in the Internet as previously existed on Minitel, the process is sometimes uncertain.

One of the interesting issues for the digital environment is the opportunities for new product bundles which were not available in print form.

Back to Top

Jeff MacKie-Mason

Jeff is a professor at the University of Michigan's School of Information, which attempts to bring technology, information science, and social science to work in the areas where these disciplines interface and intersect. He also founded the Program for Re search in the Information Economy (PRIE) <http://www.si.umich.edu/~prie/>. Some of his most recent areas include:

  • Economics of and Economics for Multiple Quality of Service Networks.
    Some questions are how to allocate resources/objects to users with different needs - how to recover fixed costs. When users have different valuations, how to extract the variable value.
  • Content Economics
    Interaction between system architecture and economics/value/incentives. (Difference between Cablet TV and Internet) Finding different ways to extract value for content, focussing on new opportunities, such as bundling and non-linear pricing for a given bu ndle.

Ongoing Projects

  • PEAK Project - Redistributes Elsevier articles, with experimental pricing and bundling mechanisms. Using a generalized subscription - prepay without identifying specific articles.
  • MARX Project - Designing mechanisms, infrastructure, and system tools for facilitating information exchange.

Back to Top

Christine Vanoirbeek

Christine works in the area of document-based technology development. She studies the transmission and use of documents. In the document-model, the added value of the electronic format (what can be done electronically which was impossible in the print wor ld?) is an important consideration. Are e-journals merely digitized versions of the print journals, or are they fundamentally different. Some of the areas under consideration are methods for exchanging and annotating documents, handling heterogeneous form ats, creating an environment for information production and dissemination. Some of the general goals are to develop methods for restructuring and adding structure to documents; to accomplish this, there needs to be analysis at various levels: document lev el, syntactic level, linguistic level, and statistical level.

Some of the work in building applications has been done in the education domain. The design of tools to deliver interactive courses over the Internet has explored some of the issues above. The communication, here, is between the professor, teaching assist ants, and students. Some of the difficulties have arisen when trying to express complicated communication, such as formulas.

Back to Top

Jakka Sairamesh

General Interests: Applied Economics, particularly pricing in Information Networks. Developing economic models of information networks, with varying architectures, charging structures, usage patterns. Many of these ideas apply to digital libraries: for ex ample work has been done on the NCESTRL project. One of the issues which arises in a digital library context is the metadata requirements for developing a useful economic model for the library. In his IBM work, he has investigated large-scale networks, lo oking at the possible architectures for such networks.

One other interest is that of the psychology/ecology of information systems. How do players interact, in terms of the cost, quality and profit of information services, and which services can be supported?

Back to Top

Second Day's Presentations (Research Agendas)

Ann Okerson

Research Priorities:

  1. Study information system use, usage, and users, obtaining some sense of value. We need to really specify what we need from information systems.
  2. Licensing agreements - ILL of e-journals (JMM's PEAK project investigates some of this area. Libraries need data to help them make rational decisions.
  3. Co-development consortiums (groups of libraries negotiating for groups of users). Is this effective in a digital library?
  4. Standard licensing language

Back to Top

Costis Dallas

Research Priorities:

At the moment we are focussing on the use of digital libraries for scholarly communication, which may be appropriate, but we should recall that there are, of course, many other potential uses as well.

  1. Information Economics at a macro level as well as a micro level
  2. Need to Match functionality of existing research academic libraries and digital libraries. Don't throw away the value of research libraries, such as guaranteed access to basic services, selection of stock, indexed store of knowledge.
  3. Development/Study of business models and organizational levels

Back to Top

Bernard Rous

Research Priorities:

  1. To overcome heterogeneity problem, look at aggregators, distributed systems, and interoperability, rather than trying to produce all the information in one system.
  2. Review intellectual property law in view of current technology.
  3. What do publications actually cost? Compare print with digital, and institutional with self-organized.
  4. Sociology of scholarship. How is the literature being used? What is changing as authors becomes both more networked and more specialized?

Back to Top

P. Bernt Hugenholtz

Research Priorities:

  1. Re-evaluate the appropriateness of the intellectual property laws
  2. Determine how digital libraries can function legally under copyright
  3. Exemptions to copyright (e.g. fair use) are important, but should be reviewed, looking specifically at the rationale. Those motivated by the preservation of user freedoms should be kept. Those motivated by social and cultural goals may need to be rede fined. Those motivated by a fear of market failure should be abolished.
  4. In redefining social/cultural - based exemptions, need to look at functional exemptions
  5. Contractual solutions. Analyze roles in information chain; analyze relevant rights and limitations; develop standard contracts; preserve user freedoms.

Back to Top

Pam Samuelson

Research Priorities:

(Generally in agreement with Bernt)

  1. Technological means for protecting copyrighted works have unanticipated consequences. While they protect fair use and public domain, they may violate privacy. Need to investigate the requirements of technology - to what extent should policy decisions be incorporated in the code itself. If builders of digital libraries don't take public policy into account, it will, eventually, be a legal problem.

Back to Top

Sebastien Steinmetz

Research Priorities:

  1. Cost structure of e-publications. Publisher and consumer costs
  2. New economic issues: bundling and non-linear pricing.
  3. Investigate the more interactive nature of digital libraries. Does producer or consumer innovate with bundling options? How can transactions now take place (e.g. auctions)?
  4. Look at computational limitations. To what extent does computer-interaction effect bounded rationality of people?

Back to Top

Jeff MacKie-Mason

Slide Presentation (Power Point file)

Back to Top

Christine Vanoirbeek

Research Priorities:

  1. Comprehension of digital libraries area. Develop a generic framework. Determine impact of new information and communication technology. Investigate business models.
  2. Contribution of document technology/document engineering discipline. Enhancing quality of service, in terms of standardization, searching, indexing, hypertext, and new services.

Back to Top

Jakka Sairamesh

Research Priorities:

  1. Determine which economic models we should be considering
  2. Understand costs in interaction between users, systems, brokers. This information will help decide architecture.

Back to Top

Presentations from Chairs

Christos Nikolaou

Current Projects:

Coastal Zone Management Project: Large amounts of heterogeneous information, such as experimental data, statistics, documents, images, video. How does the system meet the users' needs? User want to combine mathematical models with experimental data, which involves matching mathematical assumptions to experimental conditions. There are large metadata/search and retrieval issues here.

Some other common issues which have arisen in this project include mediation through broker or gateway services, which has metadata implications, and use of generic middleware, which requires the definition of digital library objects. Long this line, Chri stos' group has developed a scripting language to build broker or helper agents, which can manage dynamic workflows, and may span several machines. They have also been working on the scheduling and planning issues surrounding these agents. How do they mee t the search and user constraints, with respect to content, cost, and access? Another effort has been to develop measures of performance of distributed search mechanisms, which can be used in load-balancing schemes.

Interests:

  1. How much can market mechanisms influence the external world and the behavior within the system itself (and are the two related). For example, does the allocation of bandwidth for desired quality of service (ref. To JMM's work) effect external world as well as the system itself?
  2. Layers of economies. What are the interactions; how do they interact; are they unstable; how do we avoid unstable economies?

Back to Top

Michael Wellman

Download Slide Presentation (PowerPoint file)

Current Projects:

  1. University of Michigan Digital Library (UMDL). Building architecture to support changing boundaries between information producers and consumers. Using market-based architecture (extreme commodification of information) to introduce new services and opp ortunities more quickly. Two of the components of this architecture are the Service Classifier, which dynamically integrates new services into the system, and the AuctionBot, which uses auctions as a general market and pricing mechanism. Some of the open issues are the enforcing of legal issues, and multi-parameter negotiation, beyond price and quantity.
  2. Journal of Artificial Intelligence Research (JAIR). E-journal, which Michael is editing. The journal started because of dissatisfaction with the status quo. The journal production is a distributed, digital process from beginning to end. Functions such as in-depth copy editing are the responsibility of the author, rather than the journal. At the moment, both revenues and costs are negligible, but both may increase in the future.

Research Agenda:

  1. Economic/Payment issues: Authentication, validating object/watermarking, entitlement reasoning.
  2. Digital Publishing issues: innovative services, market research, archiving, formats.

Back to Top

Closing Discussion

Several potential research questions were generated by group members, including:

  1. Study- how to reconceptualize in/for a network information environment.
  2. Study - cost of making and delivering "first copy" of digital resources in a networked environment. (variable format)
  3. Study - how to measure (specify measures) use of digital information and use data to help determine "value" of information.
  4. Study - changing needs, behaviors, of information creators and users in a networked environment.
  5. Build- basic building blocks to experiment with models and "what if" questions.

Jeff's Framework

Structures various research areas into three categories: services, architecture and mechanisms, and policy issues.

Download Slides (PowerPoint file)

Provisional Strategy for Report Structure:

  1. Jeff's framework (or some version of it). Explanation and rationale.
  2. Further discussion of specific questions in each area. Survey landscape and point out highlights.
  3. Priority list.

Michael will plan and organize the agenda for the next meeting (Ann Arbor, MI - Spring 1998). Group members may work asynchronously before arrving at the next meeting.

Back to Top