B  U  L  L  E  T  I  N


of the American Society for Information Science and Technology       Vol. 30, No. 6      August/September  2004

Go to
Bulletin Index

bookstore2Go to the ASIST Bookstore

Copies

Our Members

    Editor's Note: In recent years the Bulletin has asked the winners of the ASIS&T Research Award to describe their careers and interests for the Bulletin. The 2003 winner was Danish information scientist Peter Ingwersen.

Highlights of a Career in Information Science
by Peter Ingwersen

Peter Ingwersen is affiliated with the Royal School of Librarianship and Information Science, Copenhagen, Denmark. He can be reached by e-mail at pi@db.dk or through his website at www.db.dk/pi

As a matter of fact, I would have liked to become either an historian or a geologist (like my father), specializing in techniques for dating artifacts and geological events, such as the Carbon 14 test. Instead I graduated in library and information studies at the Royal School of Librarianship in Copenhagen in 1973. My thesis was an empirical study using observation and interviews, as well as quantitative analyses, of the use of catalogs in public libraries in Denmark. That was prior to the online age, so essentially the thesis dealt with the patrons' use of card catalogs, including subject access, and the initial interaction with the librarians when patrons could not find their way. Looking back on these activities some 30 years ago it seems evident that already at that time I liked making quantitative empirical investigations; looking into the processes of retrieval; and digging into the historical perspectives of the objects researched.

Immediately the Royal School employed me as a lecturer in information storage and retrieval, cataloguing and indexing theory. I was very lucky because I came to work with Professor Paul Timmermann, who was trained in the research spirit of the Niels Bohr Institute (nuclear physics), that is, highly rigorous experimental settings and research approaches. Paul was extremely good at producing novel research ideas and questions, which were carried out by younger researchers – such as I. He also emphasized the international dimensions of library and information science and encouraged international publishing. So from 1976 to 1980 I was part of a research team that did experimental research on cognitive processes concerned with user-intermediary-system interaction. However, being a Danish academic at that time did not automatically mean that one could write fluently in English. Writing in English was in fact very cumbersome to begin with. But I soon got a tremendous boost.

I became a research fellow in 1982-84 at the online service of the Information Retrieval Service of the European Space Agency (ESA-IRS), located in Frascati, Italy. My R&D activities were concerned with user-system interface improvements, the development of a new family of online support and retrieval tools and online systems management. For instance, I was part of the group that invented the online term frequency feature, ZOOM, much later taken up by Dialog as the Rank command. Also, I wrote database descriptions and manuals and implemented online databases, all of which provided a lot of help in improving my written English. Most importantly, the international research environment at ESA directly supported and encouraged me to become more scientific in the sense of hard science. Most researchers came from such disciplines, and I am very grateful for the shaping and sharpening of my profile as a researcher that took place at ESA.

Back again at the Royal School of Librarianship in Copenhagen I got tenure as an associate professor in 1984. Owing to my new experiences concerned with technical and scientific information systems and systems design, obtained during the ESA period, I started up a new department dealing with IRM and design of specialized information services and systems for industry and institutions, in collaboration with my wife-to-be, Irene Wormell. During my Italian experience my research interest started to concentrate on information interaction and intermediary (interface) functionality for IR. Probably for that reason I was invited during the spring term 1987 to be a visiting professor in the School of Communication, Information and Library Studies at Rutgers University to work with Professor Nick Belkin.

Nick and I knew each other from many conferences dating back from the mid-70s – the so-called IRFIS conferences. We had been exchanging models and ideas on the cognitive viewpoint for interactive IR for more than 10 years when I came to Rutgers. At that time we also attempted to bridge the more user-centered and interactive research environment to the mainstream IR laboratory community with the help of many key scientists from both sides. This attempt was probably too advanced at that time, but it paved the way for the more interactive tracks in the NIST TREC (National Institute of Science and Technology Text Retrieval Conferences) research activities that were soon to appear. My interest in interface issues dropped after I had published my book, Information Retrieval Interaction, 1992, which was based on cognitive views and my doctoral thesis. It really seemed obvious that the expert system approach to interactive IR (IIR) was too narrow and unattainable.

Instead I became very interested in the theoretical development of IIR during the 1990s. Since our laboratory facilities were not adequate for taking part in TREC we concentrated on modeling. At the same time, around 1990, I became one of the driving forces behind the curriculum development of the master of science (Msc) program in library and information science at the Royal School of LIS and later the development of the Ph.D. program. I was appointed head of the MSc program from 1990-93. In July 1993 I switched over to research to become head of the Department of Information Retrieval Theory which, owing to a new constitution for the Royal School as a university in 1999, merged with the Department of Information Studies. Since 2001 I am a research professor in that department.

The theoretical development of IR during the 1990s was commonly a question of producing new kinds of logic or mathematical techniques guiding (and improving) IR performance. Since my theories of IR were based on a cognitive view and were not mathematical, but conceptual, emphasizing the interactive aspects of IR, my contributions on, for instance, polyrepresentation at the ACM-SIGIR conferences were constantly scheduled in the user-centered sessions such as those on interface and interaction issues. In a way it is interesting (and sad) to observe that hardly any theoretical development has taken place in IR during the last decade, as can be seen from the omission of theoretical sessions at SIGIR. Anyway, the principle of polyrepresentation assumes that actors of different cognitive nature over time make different interpretations of the same information objects. Such different interpretations could be represented by, for instance, the original author's full text, an indexer's keywords added to the document in scientific contexts, and citations or inlinks that later refer to the same document. The assumption is that the more different and frequent the representations that point to the same object, directed by an information request, the more relevant/useful that object is to that information request (or searcher, or situation, etc.).

That kind of theoretical development and ensuing hypotheses on performance improvement, the meaning of relevance, and the connection to evaluation measurements require testing. Owing to its complexity, also involving human subjects, such testing might be very difficult. Fortunately, from the mid-1990s I was lucky to get some extremely intelligent, innovative and hard working students with whom I have collaborated up to this day. Pia Borlund, Marianne Lykke Nielsen, Birger Larsen and Lennart Björneborn have been the most consistent co-workers – cutting out their own edges in LIS research – but there have been others who have contributed as well over shorter intervals. Internationally, I have also had the privilege to co-author with scientists like Peter Willett at Sheffield and Kalervo Järvelin at Tampere, Finland, and to work at the European level in close cooperation with Keith van Rijsbergen's IR Laboratory in Glasgow over the years.

Probably owing to many different factors, also unknown to me, I was fortunate to receive the Jason Farradane Award 1993 from the Institute for Information Scientists, UK, and a year later the ASIS&T Distinguished Lectureship Award for my contributions to the cognitive approach to IR. I also had the privilege of chairing the 15th ACM-SIGIR Conference on Research and Development in Information Retrieval, held in Copenhagen, June 1992, and serving as program chair in 1995 and 2000.

Gradually, from the mid-1990s, I began another research track. Aside from IIR I became interested in bibliometrics/informetrics. In particular, I became hooked on doing experiments with online bibliometrics (for example, citation analyses) with a senior colleague in Denmark, Finn H. Christensen, and Irene Wormell. We formed a Center for Informetric Studies 1996-2000.

To me, as to a few other LIS scientists in the past, like Eugene Garfield, Don Swanson, Abe Bookstein and Jean Tague, bibliometrics is very closely associated with IR. In fact, bibliometric analyses can only be carried out by means of comprehensive IR, and the use of incoming citations and outgoing bibliographic references as means to improved IR seems rather obvious. Hence, when the Web started to emerge as a natural object or platform for research I carried over the ideas from bibliometrics and traditional IR into the new field of research: webometrics. From a variety of perspectives, several other LIS scientists did the same – almost simultaneously – such as Blaise Cronin, Bar-Ilan and Ronald Rousseau, who were supplemented by new powerful approaches from people like Kleinberg (hubs and cognitive authority), Lawrence & Giles (Citeseer; use of links and citations for retrieval) or Brin & Page (Google's PageRank by inlinks and anchor texts).

Currently, I am engaged in the TAPIR Project on Text Access Potentials for Interactive IR as chair of the research team, since 2001. The project is based on the polyrepresentation assumptions to be tested and concerns the development of work-based multi-evidence access to, mainly, scientific and multimedia information objects and structured texts, including Web-based environments.

You can be sure that these last 30 years have constantly been exiting times for me!


How to Order

American Society for Information Science and Technology
8555 16th Street, Suite 850, Silver Spring, Maryland 20910, USA
Tel. 301-495-0900, Fax: 301-495-0810 | E-mail:
asis@asis.org

Copyright © 2004, American Society for Information Science and Technology