Pharmacokinetic Named Entity Recognition Benchmark

Annotated data to perform Named Entity Recognition (NER) of pharmacokinetic (PK) parameters in scientific text. Training, development and test files are released in JSONL format and store the annoated data for training and evaluating NER models. Each line in the JSONL files corresponds to an annotated sentence with the following information: text: Raw sentence spans: List of entities mentioned in the sentence, specifying: (1) the character-level boundaries and (2) the entity type of each annotation SID: Unique sentence ID metadata: Metadata with unique article, paragraph and sentence identifiers and the article section from which the sentence was extracted