Patent Domain NLP

NLP in the patent domain. Overlaps with Legal Domain NLP and Scientific Text Processing.

• Classification
  • D'hondt et al 2013 - Text Representations for Patent Classification
  • Hepburn 2018 - Universal Language Model Fine-tuning for Patent Classification
  • Niu & Cai 2019 - A Label Informative Wide & Deep Classifier for Patents and Papers
• Translation
  • 2005 Workshop on Patent Translation
  • Nanba et al 2012 - Automatic Translation of Scholarly Terms into Patent Terms Using Synonym Extraction Techniques
  • Soares et al 2020 - ParaPat: The Multi-Million Sentences Parallel Corpus of Patents Abstracts
• Retrieval
  • 2003 - Overview of Patent Retrieval Task at NTCIR-3
  • Itoh et al 2003 - Term Distillation in Patent Retrieval
  • Temnikova et al 2013 - Measuring Closure Properties of Patent Sublanguages Paper is about issues in patent search
• Keyword Extraction
  • Suzuki & Takatsuka 2016 - Extraction of Keywords of Novelties From Patent Claims
• Information Extraction and Other Analysis
  • Sheremetyeva 2003 - Natural Language Analysis of Patent Claims
  • Galibert et al 2010 - Hybrid Citation Extraction from Patents
  • Zhai et al 2019 - Improving Chemical Named Entity Recognition in Patents with Contextualized Word Embeddings
  • Nittala & Shrivastava 2020 - The WEAVE Corpus: Annotating Synthetic Chemical Procedures in Patents with Chemical Named Entities
  • Hu & Verberne 2020 - Named Entity Recognition for Chinese Biomedical Patents
• Patent Claim Construction
  • Sheremetyeva, Nirenburg & Nirenburg 1996- Generating Patent Claims from Interactive Input
  • Tonguz et al 2021 - Automating Claim Construction in Patent Applications: The CMUmine Dataset

• EuroPat: Sentence-Aligned European Patent Corpus: website 2011 paper
• ParaPat: paper Large parallel corpus of patent abstract (68M sentences total)
• CMUmine: paper dataset backup copy Patent application dataset. Contains patent claims section, used for automatic construction of patent claims. Warning: They seem to be unaware of the EuroPat dataset, as well as the large amount of prior work on NLP for patents.

• 2003 Workshop on Patent Corpus Processing
• 2005 Workshop on Patent Translation

• Legal Domain NLP Some overlap with patent domain NLP
• Scientific Text Processing Some overlap with patent domain NLP