Highlighting diverse concepts in documents

Abstract

We show the underpinnings of a method for summarizing documents: it ingests a document and automatically highlights a small set of sentences that are expected to cover the different aspects of the document. The sentences are picked using simple coverage and orthogonality criteria. We describe a novel combinatorial formulation that captures exactly the document-summarization problem, and we develop simple and efficient algorithms for solving it. We compare our algorithms with many popular document-summarization techniques via a broad set of experiments on real data. The results demonstrate that our algorithms work well in practice and give high-quality summaries.