Team Up! Cohesive Text Summarization Scoring Sentence Coalitions

Inez Okulska

Abstract: According to Aggarwal  [6] the extractive summarization is solely about scoring sentences to maximize the topical coverage and minimize redundancy, while coherence and fluency are to be considered only in the case of abstractive summaries. It is a rather strong opinion that this paper aims to argue with by introducing different functions of text summarization, the notion of text coherence and cohesion, and last but not least by offering new methods allowing for both sentence extraction and text fluency (to a certain level). This article aims at offering a model that will be as simple as possible (but no simpler, as would Einstein put that) to satisfy the goal of informative extractive summary with a certain level of cohesion determined by sentence connectives. This abstract has been generated with the algorithm proposed in the paper.