An Ontology-Based Methodology for Building and Matching Researchers' Profiles

摘要

To support potential research collaboration, we present an ontology-based methodology for identifying common research interest among researchers. The methodology uses an ontology building algorithm to build researchers' ontological profiles from publication keywords, and then an ontology matching algorithm is used to identify common research areas and degree of matching between research profiles. Our ontology matching also considers depth weights, i.e., the depth of the ontological terms within the two profiles that are matched. The idea is the terms that are located near the bottom of the ontologies should indicate specialization of researchers, and hence attention should be paid more to matching of such terms than to matching of the terms that are closer to the top of the ontologies. We present an experiment to match profiles of researchers in the same field, close fields, and different fields, and report the performance of the methodology and an evaluation using an ontology matching benchmark. The methodology is considered useful as it can quantify similarity of research interests and give practical matching results.