BlogNEER: Applying Named Entity Evolution Recognition on the Blogosphere
Paper i proceeding, 2013

The introduction of Social Media allowed more people to publish texts by removing barriers that are technical but also social such as the editorial controls that exist in traditional media. The resulting language tends to be more like spoken language because people adapt their use to the medium. Since spoken language is more dynamic, more new and short lived terms are introduced also in written format on the Web. In teTahmasebi2012 we presented an unsupervised method for Named Entity Evolution Recognition (NEER) to find name changes in newspaper collections. In this paper we present BlogNEER, an extension to apply NEER on blog data. The language used in blogs is often closer to spoken language than to language used in traditional media. BlogNEER introduces a novel semantic filtering method that makes use of Semantic Web resources (i.e., DBpedia) to gain more information about terms. We present the approach of BlogNEER and initial results that show the potentials of the approach.

Named Entity Evolution

Semantic Web

DBpedia

Blogs

Författare

Helge Holzmann

L3S Research Center

Nina Tahmasebi

Chalmers, Data- och informationsteknik, Datavetenskap

Thomas Risse

L3S Research Center

3rd International Workshop on Semantic Digital Archives, SDA 2013 - Co-located with 17th International Conference on Theory and Practice of Digital Libraries, TPDL 2013; Valetta; Malta; 26 September 2013

1613-0073 (ISSN)

Vol. 1091 28-39

Styrkeområden

Informations- och kommunikationsteknik

Ämneskategorier

Språkteknologi (språkvetenskaplig databehandling)

Mer information

Skapat

2017-10-07