Multi-document summarization is a process of automatic generation of acompressed version of the given collection of documents. Recently, thegraph-based models and ranking algorithms have been actively investigated bythe extractive document summarization community. While most work to datefocuses on homogeneous connecteness of sentences and heterogeneous connectenessof documents and sentences (e.g. sentence similarity weighted by documentimportance), in this paper we present a novel 3-layered graph model thatemphasizes not only sentence and document level relations but also theinfluence of under sentence level relations (e.g. a part of sentencesimilarity).
展开▼