Entity relatedness for retrospective analyses of global events

Nanni, Federico ; Ponzetto, Simone Paolo ; Dietz, Laura

PDF (Entity Relatedness for Retrospective Analyses of Global Events)
web_sci_16.pdf - Accepted

Download (367kB)

URL: https://ub-madoc.bib.uni-mannheim.de/40914
Additional URL: https://www.semanticscholar.org/paper/Entity-Relat...
URN: urn:nbn:de:bsz:180-madoc-409142
Document Type: Conference or workshop publication
Year of publication: 2016
Book title: NLP+CSS workshop at Web Science 2016 : May 22, 2016, Hannover, Germany
Page range: 1-5
Conference title: NLP+CSS workshop at Web Science 2016
Location of the conference venue: Hannover, Germany
Date of the conference: May 22, 2016
Place of publication: New York, NY
Publishing house: ACM
Related URLs:
Publication language: English
Institution: School of Business Informatics and Mathematics > Wirtschaftsinformatik III (Ponzetto 2016-)
Subject: 004 Computer science, internet
Abstract: Tracking global events through time would ease many diachronic analyses which are currently carried out manually by social scientists. While entity linking algorithms can be adapted to track events that go by a common name, such a name is often not established in early stages leading up to the event. This study evaluates the utility of entity relatedness for the task of identifying related entities and textual resources that describe the involvement of the entity in the event. In a small study we find that simple relatedness methods obtain MAP score of 0.74 outperforming many advanced baseline systems such as Stics and Wiki2Vec. A small adaptation of this method provides sufficient explanations of entity involvement or 68% of relevant entities.
Additional information: Nicht im WebSci '16 Proceedings-Band : http://dl.acm.org/citation.cfm?id=2908131

Das Dokument wird vom Publikationsserver der Universitätsbibliothek Mannheim bereitgestellt.

Dieser Datensatz wurde noch nicht in die Unibibliographie aufgenommen, es ist eine Unveröffentlichte Publikation. Diese Publikation als "Jetzt erschienen" melden.

Metadata export


+ Search Authors in

+ Download Statistics

Downloads per month over past year

View more statistics

Actions (login required)

Show item Show item