Summarization, simplification, and generation: The case of patents
作者：
Highlights：
• Patents contain valuable technical information.
• We present approaches that summarize, simplify, or generate sections of a patent.
• Patents’ language makes them hard yet interesting for state-of-the-art NLP tools.
• Patent summarization has moved from extractive to neural abstractive approaches.
• Simplified patents are more accessible for lay people and experts.
摘要
•Patents contain valuable technical information.•We present approaches that summarize, simplify, or generate sections of a patent.•Patents’ language makes them hard yet interesting for state-of-the-art NLP tools.•Patent summarization has moved from extractive to neural abstractive approaches.•Simplified patents are more accessible for lay people and experts.
论文关键词：Natural Language Processing,Patent mining,Summarization,Simplification,Natural Language Generation,Survey
论文评审过程：Received 29 March 2021, Revised 19 August 2021, Accepted 17 May 2022, Available online 26 May 2022, Version of Record 7 June 2022.
论文官网地址：https://doi.org/10.1016/j.eswa.2022.117627
