Contextual Networks and Unsupervised Ranking of Sentences

TL;DR

This paper introduces CNATAR, an unsupervised method using contextual networks to rank sentences, outperforming existing models and human judges on multiple datasets by leveraging syntactic and semantic relations.

Contribution

The paper presents CNATAR, a novel unsupervised algorithm that constructs contextual networks for sentence ranking, achieving state-of-the-art results across several benchmark datasets.

Findings

CNATAR outperforms human judges on SummBank dataset.

It achieves the highest ROUGE scores on DUC-02.

CNATAR surpasses previous supervised models on CNN/DailyMail and NYT datasets.

Abstract

We construct a contextual network to represent a document with syntactic and semantic relations between word-sentence pairs, based on which we devise an unsupervised algorithm called CNATAR (Contextual Network And Text Analysis Rank) to score sentences, and rank them through a bi-objective 0-1 knapsack maximization problem over topic analysis and sentence scores. We show that CNATAR outperforms the combined ranking of the three human judges provided on the SummBank dataset under both ROUGE and BLEU metrics, which in term significantly outperforms each individual judge's ranking. Moreover, CNATAR produces so far the highest ROUGE scores over DUC-02, and outperforms previous supervised algorithms on the CNN/DailyMail and NYT datasets. We also compare the performance of CNATAR and the latest supervised neural-network summarization models and compute oracle results.