References & Citations
Computer Science > Computation and Language
Title: Unsupervised Summarization for Chat Logs with Topic-Oriented Ranking and Context-Aware Auto-Encoders
(Submitted on 14 Dec 2020 (v1), last revised 25 Jun 2021 (this version, v2))
Abstract: Automatic chat summarization can help people quickly grasp important information from numerous chat messages. Unlike conventional documents, chat logs usually have fragmented and evolving topics. In addition, these logs contain a quantity of elliptical and interrogative sentences, which make the chat summarization highly context dependent. In this work, we propose a novel unsupervised framework called RankAE to perform chat summarization without employing manually labeled data. RankAE consists of a topic-oriented ranking strategy that selects topic utterances according to centrality and diversity simultaneously, as well as a denoising auto-encoder that is carefully designed to generate succinct but context-informative summaries based on the selected utterances. To evaluate the proposed method, we collect a large-scale dataset of chat logs from a customer service environment and build an annotated set only for model evaluation. Experimental results show that RankAE significantly outperforms other unsupervised methods and is able to generate high-quality summaries in terms of relevance and topic coverage.
Submission history
From: Yicheng Zou [view email][v1] Mon, 14 Dec 2020 07:31:17 GMT (259kb,D)
[v2] Fri, 25 Jun 2021 07:39:46 GMT (259kb,D)
Link back to: arXiv, form interface, contact.