Detecting Extended Contexts in Tweets using DBpedia

In an era of extensive social media usage, we have enormous amount of user data generated everyday. Finding relevant information from Twitter amongst a constant stream of messages is a herculean task considering the complex sentimental nature of data. Multiple applications across the world now rely...

Full description

Saved in:
Bibliographic Details
Published in2019 IEEE Region 10 Symposium (TENSYMP) pp. 151 - 156
Main Authors Venkatesha, M, Rao, Prasanth G, Kanavalli, Anita, Shenoy, P Deepa, Venugopal, K R
Format Conference Proceeding
LanguageEnglish
Published IEEE 01.06.2019
Subjects
Context Modeling
DBpedia
Extended Contexts
Ontology
Spotlight
Twitter Data Mining
Online AccessGet full text

Cover

More Information
Summary:In an era of extensive social media usage, we have enormous amount of user data generated everyday. Finding relevant information from Twitter amongst a constant stream of messages is a herculean task considering the complex sentimental nature of data. Multiple applications across the world now rely on Twitter for various domain sensitive and analytical use-cases. This paper proposes a scalable technique which can be used for a better context modeling of tweets generating primary and extended contexts for a set of tweets. We use ontologies from DBPedia as the resource for generating the contexts and subsequently find relevant extended contexts. DBpedia Spotlight in conjunction with DBpedia Ontology is the backbone for this proposed model. Using tweets from a twitter trend as the input, the framework output is validated against the similarity scores between generated contexts. Assuming the trend is confined to a bounded context, similarity scores calculated shows the advantages of this new approach.
ISSN:2642-6102
DOI:10.1109/TENSYMP46218.2019.8971256