Limits...
BioCreative V track 4: a shared task for the extraction of causal network information using the Biological Expression Language.

Rinaldi F, Ellendorff TR, Madan S, Clematide S, van der Lek A, Mevissen T, Fluck J - Database (Oxford) (2016)

Bottom Line: Given the complexity of the task, we designed an evaluation methodology which gives credit to partially correct statements.We identified various levels of information expressed by BEL statements, such as entities, functions, relations, and introduced an evaluation framework which rewards systems capable of delivering useful BEL fragments at each of these levels.The aim of this evaluation method is to help identify the characteristics of the systems which, if combined, would be most useful for achieving the overall goal of automatically constructing causal biological networks from text.

View Article: PubMed Central - PubMed

Affiliation: Institute of Computational Linguistics, University of Zurich, Zurich, Switzerland fabio.rinaldi@uzh.ch juliane.fluck@scai.fraunhofer.de.

No MeSH data available.


Training data example.
© Copyright Policy - creative-commons
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4940434&req=5

baw067-F2: Training data example.

Mentions: Publication references are provided as supporting information for each statement. Most BEL statements represent relationships between one BEL term and another BEL term or a subordinate BEL statement. Example BEL statements related to an evidence sentence are shown in Figure 2. The statements typically encode a semantic triple (subject, predicate and object). The predicate is one of the BEL relationship types describing the relationship between the subject and object. For track 4, we selected in particular causal relationships as shown in Table 1.Figure 2


BioCreative V track 4: a shared task for the extraction of causal network information using the Biological Expression Language.

Rinaldi F, Ellendorff TR, Madan S, Clematide S, van der Lek A, Mevissen T, Fluck J - Database (Oxford) (2016)

Training data example.
© Copyright Policy - creative-commons
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4940434&req=5

baw067-F2: Training data example.
Mentions: Publication references are provided as supporting information for each statement. Most BEL statements represent relationships between one BEL term and another BEL term or a subordinate BEL statement. Example BEL statements related to an evidence sentence are shown in Figure 2. The statements typically encode a semantic triple (subject, predicate and object). The predicate is one of the BEL relationship types describing the relationship between the subject and object. For track 4, we selected in particular causal relationships as shown in Table 1.Figure 2

Bottom Line: Given the complexity of the task, we designed an evaluation methodology which gives credit to partially correct statements.We identified various levels of information expressed by BEL statements, such as entities, functions, relations, and introduced an evaluation framework which rewards systems capable of delivering useful BEL fragments at each of these levels.The aim of this evaluation method is to help identify the characteristics of the systems which, if combined, would be most useful for achieving the overall goal of automatically constructing causal biological networks from text.

View Article: PubMed Central - PubMed

Affiliation: Institute of Computational Linguistics, University of Zurich, Zurich, Switzerland fabio.rinaldi@uzh.ch juliane.fluck@scai.fraunhofer.de.

No MeSH data available.