B U L L E T I N
Annual Meeting Coverage
The Future in Person: A Student Member's View of ASIS&T 2003
Caryn L. Anderson is a master's candidate in the Graduate School of Library and Information Science at Simmons College in Boston, Massachusetts. She is an associate with Lee Mizell Consulting, conducting research, analysis, evaluation and strategic planning for government and non-profit organizations. Additionally, she provides international accounting services to The Christian Science Publishing Society. She can best be reached via e-mail: CarynLAnderson@yahoo.com.
A student member's view of her first ASIS&T conference may not carry the weight of decades of experience, but there is still a place for the vision of those too new to the field to believe there are things that cannot be done.
This essay outlines a coordinated view of some of the technologies, theories and methods presented at the Annual Meeting of the American Society for Information Science and Technology (ASIS&T) in Long Beach, California, during the week of October 19-22, 2003. This view reflects perhaps 25% of the ideas presented. Such, alas, is the effect of concurrent sessions. My interest lies in assisting those who are facilitating complex processes ranging from city planning and industrial ecology to global conflict negotiation, public health and international development. I have included those ideas and sessions that seem to contribute to the dream of a system for integrating research from different fields – a system that will effectively serve the multi-disciplinary information needs of these urgent and complex problems.
This review is not the first articulation of such a dream, nor is it the first discussion of the state of the library and information science field in relation to this vision. Federated research integration systems or portals already exist in developing forms ranging from programs like MetaLib, an integrated library metasearching system by Ex Libris and operating at Boston College as Metaquest (http://metaquest.bc.edu/V) to simple "my library" customization options. However, for those interested in improving cross-disciplinary communication and collaboration, there was much reason for cheer at ASIS&T 2003.
Though they may not have specifically discussed it when presenting their work, many of the researchers and practitioners were effectively building critical components of super-systems. The systems in sight would be networked and customized cross-disciplinary research services and resource/citation management programs that will enable researchers, practitioners and policy makers to monitor and access monographs, journals and data set repositories integrated within an active network of peers sharing theories, methods and/or content through a fully indexed, interactive and continuous peer review process for discovering and developing knowledge.
The following four sections outline how research presented at (or near) ASIS&T 2003 will contribute to the ongoing development of these systems in the months and years ahead.
NISO Metasearch Initiative. At its business meeting held at ASIS&T 2003, the National Information Standards Organization (NISO) reported on its recently established Metasearch Initiative to develop standards for single system searching across multiple databases and electronic resources. If the languages for inter-system communication can be agreed upon, then, theoretically, all catalogs can find and access all resources. The key issues of the Metasearch Initiative include
Developments in standards will be instrumental in facilitating growth in metasearching technologies.
Domain Analysis: In 1971, the United Nations Educational, Scientific and Cultural Organization (UNESCO) and the International Council of Scientific Unions (ICSU) developed a model of scientific and technical communication called the UNISIST model. This model traces the flow of information from knowledge producer to user through the stages of publication, review, storage, indexing and retrieval. Trine Fjordback Søndergaard, Jack Andersen and Birger Hjørland of the Royal School of Library and Information Science in Copenhagen, Denmark, have updated this model to represent modern methods of electronic communication and to establish a framework for analyzing different domains.
The updated work draws on the well known characterization by psychologist Anthony Biglan in 1973 of the differences among academic areas and the way scholars interact, conduct knowledge development activities and share/publish their knowledge. Since then many studies indicate that disciplines communicate in different ways. Generalizations have been made about the social sciences, or "soft" (per Biglan) disciplines, that scholars in these fields communicate more through monographs than journals; rely more highly on the authority of older work in their field, and prefer to research via citation chaining. Conversely, scholars in the "hard" (per Biglan), or exact sciences, appear to communicate more through periodical publications; prefer the latest research and search for it directly in indexing databases.
The updated UNISIST model addresses these three communication variables and provides a means by which to incorporate what Hjørland (see related article in this issue of the Bulletin) recognizes as key epistemological differences between and within disciplines that affect how scholars communicate. Differences in philosophical approaches to knowledge development affect the structure of papers and the elements that constitute effective argument, valid research or even the definition of the problem being investigated. These distinctions in turn affect the way access points to those documents are established. A more thorough understanding of them is being developed through Hjørland's 11 approaches to domain analysis and the updated UNISIST model of Fjordback Søndergaard, et al (both published separately in the Journal of Documentation). As domain analysis develops, epistemological and communication preferences can be used as tools for addressing the NISO key metasearching issue "resource description" to assist users in selecting the appropriate resources for their search.
Thesaurus Development. The October/November 2003 ASIS&T Bulletin included an article on the new Biocomplexity Thesaurus launched by the National Biological Information Infrastructure (NBII) and CSA (Cambridge Scientific Abstracts). The new thesaurus merges six different thesauri ranging in special subjects from pollution to sociology. While this project was not presented at ASIS&T 2003, it illustrates the growth of structures available to support cross-disciplinary searching.
International Interoperability. Gregory Shreve at Kent State University's Institute for Applied Linguistics discussed the ability of existing standards to accommodate the needs of multilingual searching. While much of his work focuses on different words with the same meanings, he is equally concerned with localization and style issues. It is not just that different languages use different words but that speakers also have different writing styles and respond differently to icons. Shreve recommends developing a referential, concept-based, multilingual glossary rather than attempting to include dictionary-based "equivalents" in cataloguing records. He maintains that although many systems are not designed for internationalization, existing standards like ISO 12620 can provide a framework for crafting such systems without reinventing the wheel. His work focuses primarily on internationalization, but the theoretical implications are also encouraging for translating between domains and for developing multi-disciplinary thesauri.
Knowledge Discovery Through Methods and Findings. "The goal of information retrieval systems is to identify the most precise or relevant subset of documents … by matching concepts (typically keywords) in documents with concepts in the query statement. … But biomedical researchers increasingly are seeking systems that actually extract desired facts and answer questions, not just systems that retrieve whole documents that must be read through to find the (potentially) desired information. … They need a system that allows them to quickly review retrieved results for research methods and findings …" (Revere, Proceedings, p. 52) This information is often easy to find in the figures and tables of research papers, but keyword queries do not search the captions or legends from these critical information diagrams. The folks developing the Telemakus System, however, are working to change that. Crudely speaking, their technology extracts terms from the X and Y axes of charts, and from the captions and legends, and indexes both the terms and the relationships between them (e.g., effect of caloric restriction on cancer). They then report search results according to a schema whereby surrogates are organized by research methods and findings for easy evaluation and selection. The third component of the Telemakus System is a Visual Exploration Interface that displays a map of findings (looking like a web showing links/relationships among various elements and variables studied within a particular field). This map reveals not only what is known in an area (like neoplasm research), but also "through gaps in the map, what is yet to be discovered." As other domains/disciplines are analyzed and understood, similar schema and mapping systems can be developed. Once such systems are developed, it is possible that parallel research methodology could be identified across disciplines that could generate new means of knowledge discovery and development.
Research by Attraction Rather than "Searching." Gregory Leazer of UCLA presented a model for measuring the relationship between films as a sort of inverse of the Kevin Bacon factor. Rather than measuring the distance between two actors based on what other actors they have worked with, Leazer, along with Jonathan Furner and Rachel Napper, measured the distance between films based on the combinations of actors, directors and producers involved in each film. While it may sound like statistical horseplay at first, Leazer was actually able to develop algorithms used to design a film recommendation system that appears to be more effective than those currently offered by consumer services like Netflix. Tell the system which films you like, and, by comparing the cast and production staff of all other films, the system will produce a list of recommendations for you.
Imagine this methodology applied to research papers and other scholarly material indexed via the conceptual relationship mapping technology being developed by Telemakus. Imagine incorporating a multi-disciplinary thesaurus and comprehensive domain analysis of scholarly communication and epistemological frameworks. Imagine a "research recommendation system" where relevant research (based on findings and/or methodology) could be pro-actively and automatically identified based on your profile of selected research and resources. As you add new research to your resource list, the system would continually and dynamically adjust its algorithms based on the conceptual relationships within the papers in your profile so that newly relevant research could regularly be identified and recommended for you. Your imagining may be reality before long.
Personal Digital Libraries/Portals. Dongming Zhang of Johns Hopkins University presented his studies of usage data from the Welch Medical Library "My Welch" online personal library system. The system allows users to access a variety of personally selected resources (journals, databases, e-books, etc.), receive alerts about new resources, link to personally selected websites, search PubMed, review the latest health news, keep personal notes, review tables of contents of popular journals and ask a librarian for help – all from a single, easy to use Web page. This existing system presents a useful interface for managing resources and facilitating metasearching and collaboration. Imagine a combination of "My Welch" and Boston College's "Metaquest" with the features of a "research recommendation system" as outlined above. (Explore "My Welch" at https://mywelch.welch.jhmi.edu/login/index.cfm.)
Documentation of Knowledge. The updated UNISIST model of scholarly communication presented by Fjordback Søndergaard, Andersen and Hjørland identifies a whole branch of unpublished information flow which includes pre-print databases of theses and research reports and data set repositories housed and mediated usually by scientific and research organization servers. Melissa H. Cragin at the University of Illinois is specifically studying how researchers document their experiments whose data ends up in a shared database or repository. She is interested in both the standardized metadata that accompanies the data and the narrative material that is added. The question guiding this work is "How do interdisciplinary scientists in the fields of biodiversity and neuroscience find and use information?" Cragin and colleagues Carole Palmer and Timothy Hogan presented a poster at ASIS&T 2003 entitled "Information and Discovery in Neuroscience." In addition, her article "Toward Integrative Science: Organizing Biodiversity and Neuroscience Data" in the October/November 2003 ASIS&T Bulletin is an outstanding introduction to the challenges of integrating data sets (often documented according to different standards) in order to facilitate knowledge discovery in interdisciplinary fields like biodiversity where, for example, "data collected for very different purposes, like rainfall data, will need to be included in retrieval systems for analysis along with species data." Don't let the title scare you. The article is very accessible and useful in outlining information challenges to be addressed in all disciplines.
Other Resources for Integration. While no ASIS&T Annual Meeting presentation directly addressed the inclusion of case databases, legal instruments and grant/funding opportunities with traditional information retrieval and resource management activities, it is expected that integration of these critical components will become a part of any optimal metasearch system.
Dynamic Peer Review
Dynamic Peer Review (NeoRef). Bradley Hemminger of the University of North Carolina spoke specifically about the NeoRef Open Archive System and the potential for transforming the scholarly communication process from a limited, one-shot deal into a continuum of interactive review, criticism and response. Hemminger outlined the cycle of a work in eight stages: 1) idea; 2) present to colleagues; 3) present at conference; 4) submit to journal; 5) referees demand revision for journal; 6) journal publication of final version; 7) revision to update analysis; and 8) revision with new information. Now, we only get #6, the formal journal publication. The goal of open archive systems like NeoRef is to encourage scholarly communication as a continuous process throughout each of these stages. By posting work to an archive and designing structures for peer review, criticism and response, the timeframe between ideas and developments can be greatly reduced, while useful feedback is increased. Hemminger discussed the challenges of cataloguing such voluminous communication, but it is not beyond the imagination to conceive of systems of self-indexing of responses and reviews.
Open Archives Initiative (OAI). The OAI did not present at ASIS&T 2003, but remained just under the surface of many sessions. In working to develop and promote interoperability standards, this group is focused on enabling freer and more efficient scholarly communication. This effort is partly a response to the increasingly prohibitive costs of print and subscription e-journals from private publishers. Mentioned by many presenters, OAI developments will be making large contributions in the months and years ahead.
Public Library of Science. Perhaps it is not a coincidence that the first free, online, peer-reviewed scholarly journal, Biology, debuted at the Public Library of Science (PLoS) in the same month as the ASIS&T 2003 conference. The PLoS appears to be the beginning of the future of scholarly communication – or at least a truly earnest alternative to the domination of private publishers. As PLoS works out its kinks, it is likely that lessons learned here (technologically, politically and functionally) will serve to educate the rest of the field.
Portals as Collaborative Workspaces. Brian Detlor of the business school at McMaster University suggested in his plenary address that libraries take a page from the enterprise world's book of knowledge portals to design systems that are more robust for users. He discusses the value of offering more than just content (or access) and actually providing places to work actively with others online (within the environment of information access). Business models of knowledge management portals abound and frequently include "expert systems" in addition to "workspaces" so that employees can quickly locate fellow employees who are experts in particular areas. By developing these types of systems, libraries will not only encourage knowledge discovery and development, but also support a community environment. (See Detlor's "Fostering Robust Library Portals" under "Working Papers" at www.business.mcmaster.ca/msis/profs/detlorb/publications.htm.)
Jane Starnes, senior information specialist at the Intel Corporation Library, did not present at ASIS&T 2003, but has written an excellent and accessible article based on her presentation at the 2003 ASIS&T IA Summit about the development of Intel's knowledge portal (called Knowledge Compass) for the August/September 2003 ASIS&T Bulletin. She describes how the system works, how they developed it, what the challenges were and the insights gained from the experience.
Examples from Industry: Knowledge Portal Software. Detlor mentions a number of software companies developing these kinds of systems for enterprise (business): Verity, PeopleSoft and Plumtree. Intel runs their Knowledge Compass system on a product from AskMe, Inc.
All of this research and innovation in such a short period has combined for me, generating eager anticipation about the future and an agonizing debate about which avenue to investigate first. It is an exciting time to be a library and information science student, and I am profoundly grateful that I had the opportunity to attend ASIS&T 2003 and return to share my experience of it with my fellow students. It once seemed like an impossible dream that scholars, practitioners and policy makers could search across platforms and resources to discover relevant knowledge and develop new working relationships regardless of disciplinary, geographic or epistemological boundaries. Most certainly this utopia of scholarship remains a dream. Nevertheless, I am emboldened by the spirit of Ford Prefect to add a single word to my description of it. After my experiences at ASIS&T 2003, I can say with confidence that the dream is now only "mostly" impossible – a very big step forward!
For Further Exploration
AskMe, Inc. http://www.askmecorp.com/
From ASIS&T 2003 Proceedings
From ASIS&T Bulletins
Copyright © 2004, American Society for Information Science and Technology