Master of Science
Mercer, Robert E.
Research publications reflect advancements in the corresponding research domain. In these research publications, scientists often use citations to bolster the presented research findings and portray the improvements that come with these findings, at the same time, to make the contents more understandable to the audience by navigating the flow of information. In the science domain, a citation refers to the document from where this information originates but doesn't specify the text span that is actually being cited. A more precise reference would indicate the text being referenced. This thesis develops a framework which can create a linkage between the citing sentences from the ongoing research article and the related cited sentences from the corresponding referenced documents. This citation linkage problem has been modeled as a semantic relatedness task where given a citing sentence the framework pairs this citing sentence with each sentence from the reference document and then tries to determine which sentence pair is semantically similar and which pair is not. Construction of the citation linkage framework involves corpus creation and utilizing deep-learning models for semantic similarity measurement.
Summary for Lay Audience
Research papers refer to some other research documents to bolster the proposed ideas in the ongoing paper or when ideas from those documents are used. Some times, these references are made to help the readers to build a good background over the topics. This referencing is done by means of citation. After discussing any idea a citation is made. The span of the citation may contain only one or more than one sentence. However, for the readers, this citation sentences and the citations are nothing but a link to the referenced paper. It doesn't give any specific link or hint about which section of paragraph in the reference paper is actually being referred. So, the readers have to read the whole referenced document. Citation linkage intents to reduce this burden from the readers. The idea of citation linkage is to build a framework which given a citation sentence and the cited research document tries to detects those sentences which are being referred by the citation sentence in the ongoing paper by means of checking the semantic similarity of every sentence of the referred research document against the citation sentence
Singha Roy, Sudipta, "Investigating Citation Linkage as a Sentence Similarity Measurement Task using Deep Learning" (2020). Electronic Thesis and Dissertation Repository. 6864.