Participation in TAC KBP 2017: Cold Start TEDL and Low-resource EDL

Thumbnail Image

Links to Files

https://ebiquity.umbc.edu/paper/html/id/817/Participation-in-TAC-KBP-2017-Cold-Start-TEDL-and-Low-resource-EDL

Permanent Link

http://hdl.handle.net/11603/11600

Collections

UMBC Computer Science and Electrical Engineering Department
UMBC Faculty Collection

Author/Creator

Author/Creator ORCID

Date

Type of Work

conference paper pre-print
Text

Department

Program

Citation of Original Publication

Tim Finin, Dawn Lawrie, James Mayfield, Paul McNamee, Cash Costello, Participation in TAC KBP 2017: Cold Start TEDL and Low-resource EDL, Proceedings of the 2017 Text Analysis Workshop,2017,https://ebiquity.umbc.edu/paper/html/id/817/Participation-in-TAC-KBP-2017-Cold-Start-TEDL-and-Low-resource-EDL

Rights

This item is likely protected under Title 17 of the U.S. Copyright Law. Unless on a Creative Commons license, for uses protected by Copyright Law, contact the copyright holder or the author.

Subjects

information extraction
knowledge graph
natural language processing
Low-resource
KELVIN
UMBC Ebiquity Research Group

Abstract

The JHU HLTCOE participated in the Cold Start and the edl tasks of the 2017 Text Analysis Conference Knowledge Base Population evaluation. For our sixth year of participation in Cold Start we continued our research with the kelvin system. We submitted experimental variants that explore use of linking to Freebase across English and Chinese languages and add relations beyond those required by Cold Start. This is our third year of participation in Tri-lingual EDL and first year for the EDL Pilot for 10 Low-resource Languages. We used kelvin in the Cold Start and Tri-lingual tasks and a new custom system for the Low-resource EDL task.