Context identification of sentences in research articles: Towards developing intelligent tools for the research community

Abstract

Scientific literature is an important medium for disseminating scientific knowledge. However, in recent times, a dramatic increase in research output has resulted in challenges for the research community. An increasing need is felt for tools that exploit the full content of an article and provide insightful services with value beyond quantitative measures such as impact factors and citation counts. However, the intricacies of language and thought, and the unstructured format of research articles present challenges in providing such services. The identification of sentence contexts that encode the role of specific sentences in advancing an article's scientific argument can facilitate in developing intelligent tools for the research community. This paper describes our research work in this direction. First, we investigate the possibility of identifying contexts associated with sentences and propose a scheme of thirteen context type definitions for sentences, based on the generic rhetorical pattern found in scientific articles. We then present the results of our experiments using sequential classifiersconditional random fields -for achieving automatic context identification. We also describe our Semantic Web application developed for providing citation context based information services for the research community. Finally, we present a comparison and analysis of our results with similar studies and explain the distinct features of our application.

A Framework for Annotating ‘Related Works’ to Support Feedback to Novice Writers

“…BG-EVAL-P -We did not find evidence of other works looking for strengths in background sentences. Cited Works CW-DESC (Teufel et al, 2006b) (Angrosh et al, 2012) (Fisas et al, 2016) Teufel and Fisas have a category 'Neutral' which is directly related to our category of CW-DESC. These are used like our label for descriptions of a cited work.…”

“…Teufel et al (2006b) describes a category CoCoXY that contrasts two pieces of cited work, and highlights that this is often not annotated in the literature as most works put comparisons to author's work and a cited work together. This dis-Quality/Our Labels Related Works Comparison Background BG-DESC-NE BG-DESC-EP (Teufel, 1999) (Liakata et al, 2012) (Angrosh et al, 2012) (Fisas et al, 2015) All the related works use a label of 'Background' but they do not distinguish between those that have citation evidence or not. There are some discrepancies in what these capture to ours for e.g in Angrosh this is used for sentences that provide background or introduction.…”

“…CW-COMP (Teufel et al, 2006b) Teufel includes a category CoCoXY which contrasts two pieces of cited work as our sentence label does. CW-EVAL-P (Angrosh et al, 2012) (Fisas et al, 2016) Angrosh has two labels that represent what we capture here RWS CS and RWS. The first of these labels mentions a positive (strength) in a citation sentence and in the second a positive (strength) is mentioned but the citation is not present in that sentence.…”

“…Fisas also has this label 'CRITICISM-Strength'. Gap CW-EVAL-SC (Fisas et al, 2016) (Teufel et al, 2006b) (Angrosh et al, 2012) Our evaluation category for cited works relates directly to Tuefel's category of 'Weak' -weakness of cited approach and Fisas's 'Criticism-weakness'. Angrosh labels this as 'RWSC' -sentence noting the shortcomings in the related work citation.…”

Rhetorical Classification of Anchor Text for Citation Recommendation

“…Indeed, some experimental academic retrieval tools have tried applying them to different retrieval modes ( [18], [14], [1]), and here we explore their potential application to a deeper integration with the writing process.…”