Paper: A Hybrid Approach to Multi-document Summarization of Opinions in Reviews

ACL ID W14-4408
Title A Hybrid Approach to Multi-document Summarization of Opinions in Reviews
Venue International Conference on Natural Language Generation
Session Main Conference
Year 2014
Authors

We present a hybrid method to gener- ate summaries of product and services re- views by combining natural language gen- eration and salient sentence selection tech- niques. Our system, STARLET-H, re- ceives as input textual reviews with asso- ciated rated topics, and produces as out- put a natural language document summa- rizing the opinions expressed in the re- views. STARLET-H operates as a hybrid abstractive/extractive summarizer: using extractive summarization techniques, it se- lects salient quotes from the input reviews and embeds them into an automatically generated abstractive summary to provide evidence for, exemplify or justify posi- tive or negative opinions. We demon- strate that, compared to extractive meth- ods, summaries generated with abstractive and hybrid summarization ap...