Bibliometric Maps for Aggregated Visual Browsing in Digital Libraries Andreas Strotmann University of Alberta School of Business Edmonton, Alberta, Canada strotman@ualberta.ca Dangzhi Zhao University of Alberta School of Library and Information Studies Edmonton, Alberta, Canada dzhao@ualberta.ca ABSTRACT In this paper we describe the architecture of a visual bib- liometric browsing plug-in for the growing number of digi- tal libraries that provide cited references in their document meta-data, using a simple but effective visualization method for citation network analyses we recently introduced. Citation-based network analysis methods such as co-citation analysis have long been recognized as effective tools for gain- ing insight into the intellectual structure of a field through its literature. Visualizations of these networks can help the user get an intuitive aggregated overview of the field and the interrelationships between documents or authors, which in turn can aid query expansion, search refinement, and ex- ploratory browsing. Our design calls for a visualization of the results of a multi- variate factor analysis of a bibliometric similarity matrix cal- culated from a user’s search results and/or from documents that are closely related to them. This provides the user of a digital library with an interactive map of the literature that the user is interested in, where each visual element aggre- gates different aspects of the search result (authors and/or subfields). By helping the user see the forest for the trees (i.e., a structured visual landscape of the intellectual do- main covered by the user’s search and its bibliometric vicin- ity rather than a long list of search results), these maps and the relevant links they contain promise to provide a valuable aggregated browsing tool for digital libraries. Categories and Subject Descriptors H.3.3 [Information Storage and Retrieval]: Information Search and Retrieval—Selection process ; H.3.7 [Information Storage and Retrieval]: Digital Libraries—Systems is- sues ; H.5.4 [Information Interface and Presentation]: Hypermedia—Navigation General Terms Design Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. To copy otherwise, to republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee. SIGIR 2008 Workshop on Aggregated Search July 24, 2008, Singapore Copyright 200X ACM X-XXXXX-XX-X/XX/XX ...$5.00. Keywords bibliometric information retrieval, visual browsing, aggre- gated search, search refinement, citation indexing, citation analysis, factor analysis, open system 1. BIBLIOMETRIC INFORMATION RETRIEVAL SYSTEMS Citation indexes such as SCI, Scopus, Google Scholar, and CiteSeer play two important roles: they are both a unique information retrieval system and a data source for citation study of science and technology. Results from citation anal- ysis studies can not only help understand scholarly commu- nication structures and processes but also aid information retrieval. For example, evaluative citation analysis results can help retrieve high quality documents and publications by core players (authors, institutes, countries, etc.), and re- lational citation analysis results can help expand queries through resulting clusters of documents, authors and sub- areas. A growing number of literature retrieval systems, e.g., those of the Institute for Scientific Information (now Thom- son), Scopus (by Elsevier), Google Scholar, and CiteSeer, have demonstrated the value of incorporating citation anal- ysis results into information retrieval systems by providing such information as the number of citations each document receives, various indicators of journal quality, and links to document sets and/or authors that are related to the current search through strong cocitation or cited-by links. However, they do so largely in the form of long lists rather than visual aggregates. Bibliometric information retrieval systems seek to make full use of bibliometric techniques and results in helping solve problems currently faced by information retrieval systems [4]. Most frequently, these systems use visualization tech- niques to dynamically present concept networks produced through word analysis, or to show document or author net- works created through co-citation analysis [1, 2, 3, 4, 5]. These networks can help the user get an overview of the field and the interrelationships between concepts, documents or authors, which in turn helps query expansion and search re- finement. However, a fully “bibliometrics aware” IR system that combines evaluative and relational bibliometric analysis results in aiding searching is still not available. Although it is not difficult to understand the benefit of bibliometric analysis (especially citation analysis) results in IR systems, research on bibliometric IR systems has so far been focused on word analysis and concept networks. The