A topic-based sentence representation for extractive text summarization

Abstract

We examine the effect of probabilistic topic model-based word representations, on sentence-based extractive summarization. We formulate the task of sentence selection as a binary classification problem, and we test a variety of machine learning algorithms, exploring a range of different settings for classification and modelling. A preliminary investigation via a wide experimental evaluation on the MultiLing 2015 MSS dataset illustrates that topicbased representations can prove beneficial to the extractive summarization process, compared to a TF-IDF baseline, with Quadratic Discriminant Analysis and Gradient Boosting providing the best results for micro and macro F1 score, respectively.