RSS 2.0
Critical Assessment of Information Extraction in Biology - data sets are available from Resources/Corpora and require registration.

BioCreative I

Description [2004-01-02]

The first edition of the BioCreAtIvE challenge was carried out 2004. It consisted in two main tasks, the first task was concerned with the identification of gene mentions in text and linking protein database entries to abstracts. The second task was related to the extraction of human gene product annotations with GO terms (i.e. text passages supporting those annotations).

For a detailed description of each task refer to the BioCreAtIvE I publication section. You can also consult the (not updated) original page of the contest. The participants were provided with training data at a first stage of the contest, and after a period of time obtained the test data for which they had to submit results within a relatively short period of time (to avoid any manual result generation). Each participating team could submit up to three runs and had the possibility to provide result for any of the given tasks. After a period for evaluating the results (in case of task 2 it was carried out by the EBI-GOA team and was especially work-intensive), the assessed results were returned to each participating team.

In order to discuss and present the overall results, as well as the results and strategies implemented by each team, an evaluation workshop was held in Granada, Spain. The participating teams submitted a short workshop proceedings article, describing their systems and obtained results. At this workshop, despite the presentation of the evaluated results, participants had the opportunity to give a short presentation talk describing their systems.

Finally, after a peer-review process, a selected number of articles derived from the participating systems and the evaluation articles were published in a special issue of BMC Bioinformatics.