In support of our research projects in information retrieval, we have developed an integrated multi-process software system that shepherds research data from induction through aggregation, analysis, and presentation. We combine public-domain code libraries with our own software to provide a flexible, easily- configured modular system that exposes data online for easier collaboration. The goal is to create a single online infrastructure that allows colleagues to submit, process, analyze and visualize data, and discuss and prioritize issues through a single integrated interface. We demonstrate our system within the context of the large data set provided by the Indexer's Legacy project [1].

Our Meta-dex software suite extracts content and index text from a corpus of PDF files, and generates a meta-index that references entries across an entire domain. We provide tools to analyze the individual and integrated indexes, and visualize entries and books within the meta-index. The suite is scalable to very large data sets.

Abstract The objective of this study was to characterize the types of tissue‐centric users based on tissue use, requirements, and their job or work‐related variables at the University of Pittsburgh Medical Center (UPMC), Pittsburgh, PA. A self‐reporting questionnaire was distributed to biomedical researchers at the UPMC. Descriptive and cluster analyses were performed to identify and characterize the complex types of tissue‐based researchers. A total of 62 respondents completed the survey, and two clusters were identified based on all variables. Two distinct groups of tissue‐centric users made direct use of tissue samples for their research as well as associated information, while a third group of indirect users required only the associated information. The study shows that tissue‐centric users were composed of various types. These types were distinguished in terms of tissue use and data requirements, as well as by their work or research‐related activities.

Abstract Users' queries for visual information in American history were studied to identify the image attributes important for retrieval and the characteristics of users' queries for images. The queries were collected from 38 faculty and graduate students of American history in 1999 in a local setting. Pre‐ and post‐test questionnaires and interview were employed to gather users' requests and search terms. The Library of Congress American Memory photo archive was used to search for images. Thirty‐eight natural language statements, 185 search terms provided by the participants, and 219 descriptors indicated by the participants in relevant retrieved records were analyzed to find the distribution of subject content of users' queries. Over half of the search requests fell into the category “general/nameable needs.” It was also found that most image content was described in terms of kind of person, thing, event, or condition depending on location or time. Title, date, and subject descriptors were mentioned as appropriate representation of image subject content. The result of this study suggests the principle categories of search terms for users in American history, suggesting directions for the development of indexing tools and system design for image retrieval systems.

The TREC (Text REtrieval Conference) experiments were designed to allow large-scale laboratory testing of information retrieval techniques. As the experiments have progressed, groups within TREC have become increasingly interested in finding ways to allow user interaction without invalidating the experimental design. The development of an “interactive tract” within TREC to accommodate user interaction has required some modifications in the way the retrieval tasks is designed. In particular there is a need to simulate a realistic interactive searching task within a laboratory environment. Through successive interactive studies in TREC, the Okapi team at City University London has identified methodological issues relevant to this process. A diagnostic experiment was conducted as a follow-up to TREC searches which attempted to isolate the human and automatic contributions to query formulation and retrieval performance. © 1996 John Wiley & Sons, Inc.

This paper considers the suitability and efficiency of a highly parallel computer, the ICL Distributed Array Processor (DAP), for document clustering. Algorithms are described for the implementation of the single-pass and reallocation clustering methods on the DAP and on a conventional mainframe computer. These methods are used to classify the Cranfield, Vaswani and UKCIS document test collections. The results suggest that the parallel architecture of the DAP is not well suited to the variable-length records which characterise bibliographic data.