Datasets Comparison

Results found by the paper Technology Identification from Patent Texts: A Novel Named Entity Recognition Method (manually_checked_extractions.csv). The csv is structured as follows, the first column (index) contains the entities extracted, the second (is_tech) is a binary column identifying each entity as a technology or not. The remaining columns identify a pair of methodology/patent_ipc and indicate whether each entity was found by the specific extractor in the specific IPC.

Patent

Creative Commons Attribution 4.0 International

Results found by the paper Technology Identification from Patent Texts: A Novel Named Entity Recognition Method (manually_checked_extractions.csv). The csv is structured as follows, the first column (index) contains the entities extracted, the second (is_tech) is a binary column identifying each entity as a technology or not. The remaining columns identify a pair of methodology/patent_ipc and indicate whether each entity was found by the specific extractor in the specific IPC.

Patent

Creative Commons Attribution 4.0 International