Safeguarding Copyrighted Contents
Digital Libraries and Intellectual Property Management
CWRU's Rights Management System
Tareq M. Alrashid, James A. Barker, Brian S. Christian
Steven C. Cox, Michael W. Rabne, Elizabeth A. Slotta
Luella R. Upthegrove
Case Western Reserve University
Introduction to the Project
Since its inception in 1988, the Case Western Reserve University (CWRU) digital library project has pursued a somewhat unorthodox course. Like most contemporary, university-based projects, we have focused on a unifying theme - in our case, intellectual property (IP) management. Unlike other efforts, this goal has been examined in the context of numerous small experiments rather than as a project focused on an individual collection of materials or a single service application. This freedom to examine IP management from a number of perspectives has assisted the project team in gaining a broad understanding of the issues and problems to be addressed.
The initial concept behind CWRU's digital library effort was to examine the premise that networked delivery of full-text materials and high-quality images would provide students excellent supplemental instructional resources which could be delivered directly to their dormitory rooms. In retrospect, this premise may be self-evident; however, in 1988 the "World Wide Web" did not exist and the Internet was much more limited in application and usage. In fact, when our project began, the delivery of large, high-quality images over a wide-area network presented significant technological challenges.
To develop the underlying premise into an action plan, IBM was contacted to provide an external viewpoint on digital library issues. A committee consisting of CWRU personnel (faculty, library and information services personnel) and an IBM project representative (experienced in the use and application of IBM's image management systems) worked during 1988 and 1989 to develop an overall project plan and approach. By mid-1989, a fully developed project proposal was submitted to both CWRU and IBM management for joint consideration.
The project proposal called for a series of four instructional and research library experiments to be initially performed; these experiments would test the underlying project premise. The committee suggested that the scope of the project be expanded to embrace at least one continuous media type (digital audio was recommended) and to create instances of server-based digital library software components addressing fundamental issues such as storage management, collections management, content searching, and intellectual property rights management.
Although the proposal was modestly refined during the review process, it was largely accepted as presented and became the foundation for a joint study effort undertaken by CWRU and IBM. Under the joint study agreement, the University provided the staff and facilities to implement the project proposal while IBM provided access to hardware and software resources and a project representative to act as a liaison between the project team and various research and product units within IBM.
The Initial Projects
The four library/instructional experiments initially undertaken were:
- an electronic version of a commonly used Medical School reference book;
- networked access to a Dental School radiology collection;
- a limited electronic version of the Medical School's basic sciences syllabus (essentially a custom textbook covering the first two years of study); and
- a musical scores application integrating commercial recordings with images of orchestral scores.
It was believed that these four experiments would not only effectively test the project's premise but would also be representative of other, future applications of digital library technology across the campus.
Although the details and technical challenges of each of these experiments are interesting, it is the commonality of intellectual property issues across the applications that is most striking. Typical of many multimedia applications, each of our experimental applications required the assembly of a variety of source materials.
In some instances, basic source materials could be readily integrated into a multimedia presentation because of their public domain status. Other important source materials, such as faculty prepared notes or recently published texts, were protected by copyright. Still other materials, patient records in particular, were subject to privacy considerations. Although our experiments often required only small excerpts from these protected source materials, we were, nonetheless, obligated to seek prior permission before incorporating these materials into our experimental projects.
In some cases, multimedia presentations may be quickly assembled; however, our experience demonstrated that a substantial investment of time was needed to develop a complete unit of study or a representative sample collection. In the case of our instructional experiments, for example, time consuming development tasks included: designing and developing the pedagogical objectives and methods to be used; acquisition or development of suitable authoring, management, delivery, and presentation software; and the identification, acquisition, and preparation of source materials. In light of the extended period required to create these applications, traditional educational claims of a "fair use" exemption to the copyright provisions appeared to lose credibility.
All of the original experiments, as well as many others that have been undertaken since, have been based on source materials of mixed ownership and copyright status. This mixing and matching of content resources from two or more sources has proven to be typical of higher education multimedia applications and presents special license management challenges. Our initial experiments were successfully completed using public domain and protected materials for which we received limited usage permissions. However, it was painfully apparent from the earliest days of the project that a more comprehensive system for managing IP usage permissions was needed if the goals of our project were to be realized.
Intellectual Property Management Requirements
To understand the complexity of the issues facing the CWRU project team, it is helpful to review the requirements and expectations that the providers, consumers, and purveyors of the source materials imposed on the project efforts.
Publishers and Content Providers
Regardless of the presentation medium, content owners rightfully expect formal attribution of their works and, in some cases, require compensation for its use. The mixing and matching aspects of educational use also makes it difficult for content owners to distinguish their materials from those of other contributors. Further, to the content provider, electronic distribution posed new threats and challenges to the financial viability of their products. Given that there is at least some misuse of existing published materials, they questioned how digital copies of their materials could be distributed yet protected from unauthorized and/or uncompensated usage. To this group, widespread, unattributed and/or unprotected distribution of source materials represented an unacceptable risk; they sought full identification of all users, tracking and payment for all usage, restrictions on distribution beyond campus boundaries, and high levels of security for their materials.
Students, Scholars, and Researchers
All of our experiments were directed for delivery to students, scholars, or researchers. These consumers anticipate that the University will provide network accessible materials for their use at no cost to the individual and that the University will scrupulously protect their property rights. Reference materials are assumed to be funded through library materials acquisition budgets or, in some instances, purchased by their academic department or school. Students, who already pay for their instruction in the form of tuition, object to any additional charges for use of University developed instructional materials.
Our consumers also anticipate that they will have considerable latitude in the way in which on-line materials can be used. They expect that any discrete media materials may be printed for off-line use, and also desire the ability to download or electronically cut-and-paste presentation materials into works which they create. Further, network availability enabled access whenever and wherever they desired; students, scholars, and researchers living or working off-campus and accessing institutional resources via modem expect roughly the same functionality available to on-campus users.
Faculty and Librarians
The faculty and library staff hold a unique middle ground position between the roles of supplier and the consumer. Although both groups are frequent consumers, and in many cases content authors, they also serve in an intermediary role between all other campus consumers and content providers.
Librarians typically acquire materials for use by the entire campus population. Materials obtained using library acquisition funds should, at a minimum, be available to all local library patrons. Librarians demand that the privacy and usage patterns of their patrons be thoroughly protected. Additionally, librarians provide a wide range of information support services to the entire campus community. To perform their duties they must have wide-ranging access to the entire collection of materials.
Faculty hold an especially important position with respect to acquisition and development of instructional materials. While the students ultimately purchase the required textbooks, the faculty select and adopt these materials. The faculty must have access to student materials to conduct their duties. Since the faculty frequently receive or prepare special, instructor supplements for use by all instructors, they too must have wide-ranging access capabilities and assurances that instructor materials will be secured against unauthorized distribution.
CWRU's Basic Approach
Although the requirements stated above seem incompatible, the CWRU team felt that, over time, the convenience of network accessibility would drive information consumers to more moderate positions. Similarly, we felt that content owners would moderate their requirements in the face of market demand and full comprehension of the financial benefits that a network distribution channel would offer.
We reasoned that the differences which separated the parties were largely reconcilable. We assumed that all parties would, in the end, find useful mechanisms, probably embodied within a licensing agreement, to deliver and receive digitally formatted materials over the network. We further understood that the IP management issues facing the academic community were issues that our colleagues in the commercial sector would also encounter.
If our reasoning were correct, the goal of our IP management efforts should be to encourage electronic publication and consumption of information products. To the project team, this implied a need to develop a system flexible enough to encode and reasonably enforce the terms and conditions of most licensing agreements.
In mid-1990, we began to interview faculty, librarians, attorneys, publishers, authors, trade groups and others to develop a set of property management system requirements. We supplemented this requirements set with our own "real world" experiences gained during the development of our digital library experimental applications.
Several IP management system architecture alternatives were developed and evaluated. After considering the various options, we selected an end-to-end content management approach based on the client-server model for our rights management system (RMS) development efforts. With these two decisions made, the CWRU project team created a working model of RMS during 1991 and 1992.
The RMS server was designed to act as both a content repository and as a clearing house for all client requests for access to protected materials, The server is capable of simultaneously managing and evaluating a variety of access requests and license agreements for each content item in its repository. Fully specified distribution rules, which encode the terms and conditions for delivery of IP content, are stored on the server in a series of RDBMS tables. At the server, valid requests, those that can be authenticated, result in the packaging of the following three types of information:
- applicable source materials are processed to comply with licensing requirements such as the application of encryption, "digital watermarks", or "digital fingerprints";
- copyright notices and the terms of all permissions extended to the consumer (including charge back information if appropriate); and
- an RMS trusted client application used to present the requested materials in accordance with applicable usage terms and conditions on the end-user's workstation (the RMS trusted client is dynamically inscribed on the server with a set of time dependent authentication credentials during this packaging process).
RMS defines a trusted client to be a workstation application which conforms to RMS client processing and communication conventions (RMS-compliant). An RMS client is trusted by the purveyor and content provider to exercise due diligence enforcement of all applicable licensing terms and conditions for IP delivered to the workstation. The RMS client may be trusted by the end-user to securely transit his or her network credentials over an unsecured network. Finally, the RMS client may be trusted by the system administrator to frustrate attempts by end-users to develop client applications that masquerade as an authentic RMS client application. For example, an RMS-compliant browser should neither decrypt materials which are not in active use on the client nor save unencrypted data on the client machine unless permitted under the licensing agreement. On the other hand, an RMS-compliant browser must report all usage to the server, prevent the consumer from performing any unauthorized functions (e.g., save, copy, and print), and notify the consumer in advance if they attempt to perform a function for which payment is required.
A complete discussion of the operation of an RMS client application is beyond the scope of this article; nonetheless, it would be inaccurate to portray current implementations of an RMS trusted client as completely tamper-proof. Indeed, it is difficult to imagine any software product that could today make such an unqualified claim. It is more accurate to view the RMS trusted client as one that makes tampering difficult by imposing very strict conditions on the manner and timing of credential verification. The server will only deliver materials to a client application capable of responding to credential challenges in an appropriate and timely manner.
RMS is intended for use by organizations that function as purveyors of information (publishers, on-line service providers, campus libraries, etc.). The system is capable of managing a wide variety of agreements from an unlimited number of content providers. RMS permits customization of licensing terms so that individual users, or user classes, may be granted unique privileges to specific sets of materials. As an example, agreements providing the following privileges may be easily defined:
- view-only capabilities to a textbook accessed by anonymous users located in the library,
- display/print/copy access to all on-campus students enrolled in a course for which the textbook has been adopted, and
- full access for faculty to both student- and instructor-versions of basic and supplemental materials.
Fundamental to the implementation of RMS is the creation and maintenance of a license agreement database. This database expresses the terms and conditions under which the purveyor may distribute materials to its consumers. Relevant features of RMS which the database enables include:
- A high degree of owner-defined content granularity,
- A choice of central or distributed management of the licensing databases,
- Multiple agreement types (e.g., site licensing, limited site licensing and pay-per-use), and
- Managed presentation by RMS-enabled Web browser "plug-in" modules or helper applications (trusted clients).
RMS maintains a comprehensive set of distribution permissions and rules. The premise of RMS is that each publication may be viewed as an anthology or compound document. A publication under this definition consists of one or more content elements and media types; each element may be individually managed as required. Individual content elements may be defined as broadly or narrowly as desired (i.e., the granularity of the elements is defined by the content owner); however, for overall efficiency, content elements should represent a significant and measurable unit of material. Figures, tables, illustrations, and sections of text are reasonable candidates as content elements.
To manage the distribution of complete publications or individual content elements, two licensing metaphors are implemented.
- A Master Agreement broadly defines the rules and conditions that apply to all licensing agreements negotiated between the purveyor and a content provider; only one Master Agreement may be defined between the provider and the purveyor. In practice, RMS assumes that the purveyor will enter into one or more collection agreements with each of its content suppliers. At the time the first license agreement is executed between the two parties, a Master Agreement is created along with one or more Collection Agreements (see below) in the purveyor's RMS database.
- Collection Agreements are used to specify the rules and conditions that apply to specific sets of materials licensed by a content provider to the purveyor. This agreement takes the form of a list of publications and the associated terms and conditions under which these publications may be issued to end-users (one or more Collection Agreements may be defined and simultaneously managed between the purveyor and a customer).
Individual end-users (i.e., faculty members, students or library patrons) may access the RMS-Server to search and request delivery of licensed publications. Depending upon the agreement(s) negotiated between the provider and the purveyor, individual users may be assigned access permissions based upon a combination of their user-ids and network addresses, the capabilities of their web browser, etc. Internet Protocol addresses may be used to limit distribution by physical location (e.g., to users accessing the materials from a library, a computer lab, or from an on-campus workstation). User identification may also be exploited to create limited site-licensing models or individual user agreements (e.g., distributing publications only to students enrolled in Chemistry 432 or, perhaps, to a specific faculty member).
To provide flexibility in license agreement negotiations, four distribution permissioning levels (Master Agreement, Collection Agreement, Publication, and Content-Element) may be defined. At each level, access rules, usage privileges, and usage charges may be defined within the server's RDBMS rule tables. In general, the access and usage permissions rules are broadly defined at the Master and Collection Agreement level and are refined or restricted at the Publication and Content-Element levels.
A rule may be viewed as a row within one of four rule tables. Although the specific columns within these tables vary somewhat by permissioning level, each rule expresses a condition under which access to the requested information can be granted. At each permissioning level, zero or more rules may be specified; the system administrator is free to enter as many rules as are necessary to fully define the license negotiated with the content owner. Each rule is assigned a specific start and end date consistent with the license agreement and other information such as usage type, user-id, user classification, content type, workstation location, and so forth may be entered. All rules which are in effect at the time a request for content delivery is submitted are examined to determine if the end-user and the originating workstation may be granted the access requested.
Assuming a rule is found that permits access, materials may be distributed and presented at no charge, at a royalty charge, at an overhead charge, or at a royalty plus overhead charge to the consumer. For example, a high-level Collection Agreement rule can be defined specifying that any text element within any publication governed by the license agreement may be, by default, printed at the client workstation for some fixed fee, say 10 cents per page; however, high value text sections may be individually identified with either a Publication or a Content-Element level rule and assessed an exceptional, higher charge, say 25 cents per page.
When a request for delivery of materials is received from a client machine, the server evaluates the associated content rules in a "bottom-up manner" (e.g., content-element rules are evaluated before publication rules which are, in turn, evaluated before collection agreement rules, etc.). Access and usage privileges are resolved when the system recognizes a match between the permission rules governing the content and the consumer's credentials. As noted earlier, access to content is only granted when an applicable set of rules specifically granting access permission to the consumer or client location are found. In the cases where two or more rules of similar priority permit access, the rules most favorable to the consumer are selected.
The following use of RMS database rules has been found to be effective during CWRU testing:
- Use Master rules to define the agreement's term (beginning and ending dates) and the general "fair use" guidelines (rules) negotiated between a supplier and the purveyor. This approach permits "fair use" definitions to be re-defined in response to new standards or regulatory definitions without requiring modifications to RMS itself.
- Use Collection Agreement rules to define the term (beginning and ending dates) for specific licensing agreements between the supplier and the purveyor. General access and permission rules by user-ID, user category, network address, and media type would be assigned at this level.
- Use Publication rules to impose any user-id or user category-specific rules (e.g., permissions for students enrolled in a course for which this publication has been selected as the adopted textbook) or to impose exceptions based on the publication's value.
- Use Content-Element rules to grant specific end-users or user categories access to materials (e.g., define content elements which are supplementary teaching aids for the instructor) or to impose exceptions based on media type or content value.
While RMS is designed to address all types of multimedia rights, permissions and licensing issues, the current implementation has focused on distribution of traditional print publication media (text and images) and on typical usage types (e.g., display, local print, copy, and download). Extensions to RMS are anticipated to address the distribution of a full range of multimedia types.
The project team has developed a number of the RMS-compliant browsers for use on campus. While the use of custom developed browsers is adequate for many campus applications, it is inadequate for a more general deployment of the system. Specifically, faculty and students often prefer to view content elements using commercial products with which they are already familiar. Recently, a set of RMS-compliant Adobe Acrobat plug-in modules have been developed and successfully tested. Opportunities to expand RMS services to other popular workstation applications are now being investigated.
Assessment of the Project and RMS System and Future Directions
In general, the unorthodox "many small experiments" approach adopted for the project has proven to be very beneficial. Numerous small projects have been undertaken, and each has contributed in its own way to our understanding of how to deploy digital library technologies. While not all experiments have been successful in attaining their original goals, even the failures have provided excellent lessons in how and how-not to use the technology.
Typically, experiments undertaken by the project team have not, in and of themselves, attained production status. Rather, the experiments have shown how the technology might be best used and have served as prototypes for more intensive investment in full-scale production systems.
Like any complex system, RMS does present a series of administrative challenges that must not be underestimated. The entry of provider information, specification of applicable licensing rules, and maintenance of copyright notice information can be tedious. The work of other digital library researchers, such as efforts intended to create unique digital identification codes for all content elements or to standardize the rights management language, may alleviate many of these problems. Additional developmental work to create a more intelligent administrative interface application is also under consideration. Surprisingly, rule administration has not been as difficult as we originally feared; minor variations on a set of six common rules have proven to be sufficient to manage most research journal site licensing agreements.
It has proven difficult to extend RMS services to many popular commercial browser products. These difficulties are due in large measure to basic assumptions made during the development of these products; essentially, most software developers assume that any licensed user of their software will have unlimited rights to manipulate and use the content processed by that software. It has proven to be either difficult or impossible to provide content protection in some commercial software products without the support and participation of the vendor's software development team.
The project team also recognizes that adding RMS components to commercial authoring products would be most beneficial. Since much of the descriptive data needed by RMS could be efficiently entered at the time the content is created, "plug-in" authoring modules specifically designed to capture this information and to incorporate the data into the content objects would substantially reduce subsequent administrative processing.
Overall, we have been pleased with the results of the project and with our long-term relationship with IBM. Support from others, most notably The Andrew W. Mellon Foundation, is also gratefully acknowledged. RMS itself has proven to be well received on campus and well positioned to address the needs of most content providers. The current version of RMS now forms the foundation for an increasing number of specialized digital library project development efforts on our campus.
Copyright © 1998 Tareq M. Alrashid, James A. Barker, Brian S. Christian, Steven C. Cox, Michael W. Rabne, Elizabeth A. Slotta, Luella R. Upthegrove
Top | Magazine
Search | Author Index | Title Index | Monthly Issues
Previous Story | Next Story
Comments | E-mail the Editor