Building a thailand researcher network based on a bibliographic database

Abstract

Among many practical and domain-specific tasks, expertise retrieval (ER) has recently gained increasing attention in the information retrieval and knowledge management communities. This paper describes our ongoing project to design and implement an expert retrieval system with the scope on researchers who work in In our current system prototype, we assume that the areas of expertise among researchers can be extracted from bibliographic databases. We use the Science Citation Index (SCI) database to provide the information for representing the expert profiles. From the SCI database, we queried and retrieved publications covering from the year 2001 to 2008 by specifying the affiliation equal to Thailand. The results contain a set of approximately 23,000 publications. We downloaded and extracted four related fields including authors (denoted by AU), controlled terms (denoted by ID), keywords (denoted by DE) and subject category (denoted by SC). To build a researcher network, we consider two types of relationships: direct and indirect. The direct (or social) relationship is defined as the co-authoring degree between one researcher to others. The co-authoring degree between two researchers, co-authoring(A,B), can be calculated based on the co-occurrence frequency between A and B found in the field AU of 23,000 retrieved records. The indirect (or topical relationship is defined when two researchers have publications under the same topics. The topical degree between two researchers, topical(A,B), can be calculated based on the similarity measure between two sets of extracted keywords, keyword(A) and keyword(B), representing researcher A and B, respectively. The keyword set can be extracted from the fields ID, DE and SC. An author with high frequencies on particular keywords is considered an expert in the corresponding research topics.