A Novel Feature-based Bayesian Model for Query Focused Multi-document Summarization

TL;DR

This paper introduces a new supervised Bayesian model that integrates rich sentence features with topic modeling for improved query-focused multi-document summarization, demonstrating effectiveness on TAC datasets.

Contribution

It presents a novel supervised Bayesian approach combining sentence features with topic models, advancing query-focused multi-document summarization techniques.

Findings

Effective on TAC2008 and TAC2009 datasets

Outperforms existing methods in summarization quality

Integrates features and topic models in a principled way

Abstract

Both supervised learning methods and LDA based topic model have been successfully applied in the field of query focused multi-document summarization. In this paper, we propose a novel supervised approach that can incorporate rich sentence features into Bayesian topic models in a principled way, thus taking advantages of both topic model and feature based supervised learning methods. Experiments on TAC2008 and TAC2009 demonstrate the effectiveness of our approach.