ScisummNet: A Large Annotated Corpus and Content-Impact Models for Scientific Paper Summarization with Citation Networks
This addresses the problem of limited resources and incomplete summaries for researchers in computational linguistics and related fields, though it is incremental in building on existing summarization approaches.
The paper tackled the lack of large annotated datasets and the need to include research impacts in scientific paper summarization by creating ScisummNet, a large manually-annotated corpus for computational linguistics, and proposing hybrid methods that combine author highlights and citation impacts, showing advantages over abstracts and traditional citation-based summaries.
Scientific article summarization is challenging: large, annotated corpora are not available, and the summary should ideally include the article's impacts on research community. This paper provides novel solutions to these two challenges. We 1) develop and release the first large-scale manually-annotated corpus for scientific papers (on computational linguistics) by enabling faster annotation, and 2) propose summarization methods that integrate the authors' original highlights (abstract) and the article's actual impacts on the community (citations), to create comprehensive, hybrid summaries. We conduct experiments to demonstrate the efficacy of our corpus in training data-driven models for scientific paper summarization and the advantage of our hybrid summaries over abstracts and traditional citation-based summaries. Our large annotated corpus and hybrid methods provide a new framework for scientific paper summarization research.