Difference between revisions of "Named entity recognizers"
Jump to navigation
Jump to search
| Line 7: | Line 7: | ||
* [http://l2r.cs.uiuc.edu/~cogcomp/asoftware.php?skey=NE Older version of UIUC NER] - identifies/classifies entities as Person, Location, Organization and Misc (this last category relates to languages and nationalities); fast and robust; try the [http://l2r.cs.uiuc.edu/~cogcomp/ne_demo.php demo] | * [http://l2r.cs.uiuc.edu/~cogcomp/asoftware.php?skey=NE Older version of UIUC NER] - identifies/classifies entities as Person, Location, Organization and Misc (this last category relates to languages and nationalities); fast and robust; try the [http://l2r.cs.uiuc.edu/~cogcomp/ne_demo.php demo] | ||
*[http://nlp.stanford.edu/software/CRF-NER.shtml Stanford NER] Conditional Random Fields based NER. Also incorporates distributional similarity based features extracted from the English Gigaword corpus. | *[http://nlp.stanford.edu/software/CRF-NER.shtml Stanford NER] Conditional Random Fields based NER. Also incorporates distributional similarity based features extracted from the English Gigaword corpus. | ||
| + | |||
| + | *[http://www.alias-i.com/lingpipe/ LingPipe] | ||
[[Category:Software]] | [[Category:Software]] | ||
Revision as of 20:03, 5 January 2009
Software - Named entity recognizers
- Greek named entity recognizer (version 2) - currently identifies temporal expressions, person names, and organization names; see here for publications describing the recognizer
- Balie Baseline implementation of named entity recognition.
- UIUC NER Java-based UIUC NER tagger. Uses gazetteers extracted from Wikipedia, word-class model built from unlabeled text and extensively uses non-local features. Achieves 90.8F1 score on the CoNLL03 shared task data and is robust on other datasets. Try the LBJ-NER-Demo
- Older version of UIUC NER - identifies/classifies entities as Person, Location, Organization and Misc (this last category relates to languages and nationalities); fast and robust; try the demo
- Stanford NER Conditional Random Fields based NER. Also incorporates distributional similarity based features extracted from the English Gigaword corpus.