Neural Keyphrase Generation via Reinforcement Learning with Adaptive Rewards
Refereed conference paper presented and published in conference proceedings


Full Text

Times Cited
Web of Science2WOS source URL (as at 10/06/2021) Click here for the latest count

Other information
AbstractGenerating keyphrases that summarize the main points of a document is a fundamental task in natural language processing. Although existing generative models are capable of predicting multiple keyphrases for an input document as well as determining the number of keyphrases to generate, they still suffer from the problem of generating too few keyphrases. To address this problem, we propose a reinforcement learning (RL) approach for keyphrase generation, with an adaptive reward function that encourages a model to generate both sufficient and accurate keyphrases. Furthermore, we introduce a new evaluation method that incorporates name variations of the ground-truth keyphrases using the Wikipedia knowledge base. Thus, our evaluation method can more robustly evaluate the quality of predicted keyphrases. Extensive experiments on five real-world datasets of different scales demonstrate that our RL approach consistently and significantly improves the performance of the state-of-the-art generative models with both conventional and new evaluation methods.
All Author(s) ListHou Pong Chan, Wang Chen, Lu Wang, Irwin King
Name of Conference57th Annual Meeting of the Association-for-Computational-Linguistics (ACL)
Start Date of Conference28/07/2019
End Date of Conference02/08/2019
Place of ConferenceFlorence
Country/Region of ConferenceItaly
Proceedings Title57th Annual Meeting of the Association-for-Computational-Linguistics (ACL 2019)
Book titleProceedings of the 57th Conference of the Association for Computational Linguistics, {ACL} 2019, Florence, Italy, July 28- August 2, 2019, Volume 1: Long Papers
Year2019
Pages2163 - 2174
ISBN978-1-950737-48-2
LanguagesEnglish-United States

Last updated on 2021-11-06 at 00:20