Contrastive text summarization: a survey

Abstract In our data-flooded age, an enormous amount of redundant, but also disparate textual data is collected on a daily basis wide variety topics. Much this information refers to documents related the same theme, that is, different versions document, or discussing topic. Being aware such differences turns out be important aspect for those who want perform comparative task. However, as increase in size and volume, keeping up-to-date, detecting, summarizing relevant changes between it becomes unfeasible. This motivates rise contrastive summarization task, which attempts summarize text topic way highlights them. Our research aims provide systematic literature review summarization, highlighting methods, sets, metrics, applications. Overall, we found most commonly used controversial news articles, opinions sentiments topic, reviews product. Despite great interest note standard well competitive task dedicated are yet come proposed, eventually impeding emergence new methods. Moreover, breakthrough using deep learning-based language models abstract summaries still missing.