0.25
0.5
0.75
1.25
1.5
1.75
2
Using @Twitter Conventions to Improve #LOD-Based Named Entity Disambiguation
Published on Jul 15, 20151474 Views
State-of-the-art named entity disambiguation approaches tend to perform poorly on social media content, and microblogs in particular. Tweets are processed individually and the richer, microblog-spec
Related categories
Chapter list
Using @Twitter Conventions to Improve #LOD-based Named Entity Disambiguation00:00
Named Entity Recognition and Disambiguation (NERD)00:07
NERD in Tweets00:42
We need ...02:08
New Tweet NERD Corpus02:52
YODIE03:51
YODIE Overview04:39
YODIE Scores—Structural06:29
Structural—Relatedness (Milne & Witten 2008)07:25
Structural—LOD-based similarity07:48
YODIE Scores—Text-based08:08
YODIE Disambiguation09:02
Tweet Expansions Studied09:31
GATE GUI Screenshot with Expansion Examples - 109:55
GATE GUI Screenshot with Expansion Examples - 216:57
How Expansions Make a Difference17:55
Experimental Conditions18:01
Baselines18:01
Structural Scores—LOD-based18:01
Structural Scores—Relatedness18:01
Text-based—Abstracts Only18:02
Text-based—Abstracts Plus Other Textual Fields in DBpedia18:02
Including the Disambiguation Stage18:03
Contextualizing the Result - 119:41
Contextualizing the Result - 220:22
Thanks for Listening!21:29