A reference comments crawler for assisting research paper writing

When writing a research paper, significant effort is spent comparing the current work to other related studies. In general, these comparisons comprise the ‘Related Work’ section of the paper, with the relevant reference papers cited and analyzed. An automatic method for gathering and managing information about other researchers’ reviews of reference papers would reduce the time and effort needed for such comparisons. Thus, in this paper, we propose a crawler that gathers the comments made by other researchers on the papers that are cited in the ‘Related Work’ section and listed in the ‘Bibliography’ or ‘References’ section of a research paper. The Reference Comments Crawler (RCC) system collects the text pertaining to the reference papers, providing useful information to researchers by extracting relevant data from the comments. The RCC considers different types of reference identifiers and the comment sentences are extracted based on these reference identifiers and user-defined extraction rule. Also, the RCC system extracts and provides the previous and subsequent sentences, labeled as PreSentences and PostSentences, as well as the comment sentences including the reference identifier. A series of experiments were performed to evaluate precision and recall, and the results showed that the RCC system can provide useful information with a high degree of precision and recall to the user. Furthermore, through these experiments, our system can assist researchers by reducing the time and effort spent comparing and analyzing related work.