Comparative evaluation of ontology-based Automatic Reference Tracking (ART)

Automatic Reference Tracking (ART) involves systematic and recursive tracking of bibliographic reference articles listed under the bibliography section of a particular research article. Automatic tracking is done by recursively extracting the references listed at the tail end of the input seed publication and further analysing the relevance of the extracted bibliographic reference listing with respect to the seed publication. Based on the relevance, the bibliographic research article is downloaded online. The objective is to automatically identify closely relevant reference articles, thereby facilitating the literature understanding of the aspiring researcher. In this paper, we try to address the issue of relevance-based ART in detail, with explanations on ontology-based semantic relevance computations for two domains: Operating Systems (OSs) and Computer networks (Comp n/ws). The results substantiate the claim of using domain ontology for various reasons, which are summarised in the paper.