Aggregating search results for social science by extracting and organizing research concepts and relations

A method was developed for automatically aggregating sets of research abstracts in sociology retrieved by an information re-trieval (IR) system in response to a user query. The method fo-cused on extracting and integrating research concepts, relation-ships, research methods and contextual information from different abstracts retrieved by the IR system, and organizing and display-ing them in a Web interface. Sociology dissertation abstracts were selected as the sample domain in this study. A user evaluation was carried out to evaluate the effectiveness of the aggregation method. Two concept-based presentations—with or without the use of a taxonomy—were compared against a sentence-based presentation that lists only the research-objective sentences ex-tracted from the abstracts and another sentence-based presentation generated using the MEAD system. A sample of 40 users found the concept-based presentation organized using a taxonomy to be more comprehensible, readable and useful than the sentence-based presentations. The majority of users (64%) also indicated a preference to use the concept-based system.