Skip to content

Commit

Permalink
#53 - Update documentation for 2nd release
Browse files Browse the repository at this point in the history
- added details of NER evaluation
  • Loading branch information
maxxkia committed Dec 18, 2017
1 parent 0a40f66 commit c40e3ec
Showing 1 changed file with 9 additions and 0 deletions.
9 changes: 9 additions & 0 deletions ss-doc/src/main/asciidoc/user-guide/components.adoc
Original file line number Diff line number Diff line change
Expand Up @@ -308,6 +308,15 @@ FMeasure scores
Overall F-Measure: 0.742857
----

For calculating the agreement between the predicted annotations and the gold annotations, DKPro Statistics
(https://dkpro.github.io/dkpro-statistics/) library is used. For detailed information about computation of the agreement
scores you can refer to the tutorial (https://dkpro.github.io/dkpro-statistics/inter-rater-agreement-tutorial.pdf)
provided in the DKPro Statistics project.

For calculating the precision and recall, the gold and predicted documents are converted to BIO format annotation.
Each token slot with named entity annotation (either B or I) is used to count the number of true positive, false positive, and
false negative instances.

=== Variable mention detection
With this module, you can perform variable mention detection on your data. This module can be used to detect whether
a sentence from an article contains a mention of a variable from one of the social sciences studies. The input documents
Expand Down

0 comments on commit c40e3ec

Please sign in to comment.