Corpora for Chemical Entity Recognition

Corpora for Named Entity Recognition of Chemical Compounds

The test corpus described in [Kolarik et al. 2008] is provided in the following format:

Each Entry starts with a # followed by its PMID number

The columns:

1. Token
2. Start Index
3. End Index
4. Full untokenized Entities
5. Class (B-class|I-class|O)
   • B- means: Beginning of an entity
   • I- means: Continuation of an entity
   • O means: None of the defined entities

The corpora from [Klinger et al. 2008] do not include the untokenized entities and has a differently formatted header (starting with #).

[Kolarik et al. 2008] Corinna Kolářik, Roman Klinger, Christoph M. Friedrich, Martin Hofmann-Apitius, and Juliane Fluck. Chemical Names: Terminological Resources and Corpora Annotation. In Workshop on Building and evaluating resources for biomedical text mining (6th edition of the Language Resources and Evaluation Conference), Marrakech, Morocco, 2008

[Klinger et al. 2008] Roman Klinger, Corinna Kolářik, Juliane Fluck, Martin Hofmann-Apitius, and Christoph M. Friedrich. Detection of IUPAC and IUPAC-like Chemical Names. Bioinformatics, 24(13):i268-i276, 2008.

Download

Corpus in IOB Format, gzipped, original version used in the paper [Kolarik et al., 2008]

• Original version as used for experiments described in the paper

Corpus in IOB Format, gzipped [Kolarik et al., 2008]

• Reannotated version, small differences to the version used for the paper

Corpus in IOB Format, gzipped Version 3 [Kolarik et al., 2008]

• Reannotated version in comparison to the paper, offset errors corrected

Training Corpus for IUPAC and IUPAC-like Chemical Names [Klinger et al., 2008]

• Training Corpus of 463 abstracts with 3712 IUPAC entities, 321 partial entities and 1039 MODIFIER entities.

Sampled Test Corpus for IUPAC and IUPAC-like Chemical Names [Klinger et al., 2008]

• Sampled Test Corpus of 1000 abstracts (included as full text in header of each instance). It includes 151 IUPAC entities.

Literature

[Kolarik et al. 2008] Corinna Kolářik, Roman Klinger, Christoph M. Friedrich, Martin Hofmann-Apitius, and Juliane Fluck. Chemical Names: Terminological Resources and Corpora Annotation. In Workshop on Building and evaluating resources for biomedical text mining (6th edition of the Language Resources and Evaluation Conference), Marrakech, Morocco, 2008

[Klinger et al. 2008] Roman Klinger, Corinna Kolářik, Juliane Fluck, Martin Hofmann-Apitius, and Christoph M. Friedrich. Detection of IUPAC and IUPAC-like Chemical Names. Bioinformatics, 24(13):i268-i276, 2008.