Disambiguating the Twitter Stream Entities and Enhancing the Search Operation Using DBpedia Ontology: Named Entity Disambiguation for Twitter Streams

Disambiguating the Twitter Stream Entities and Enhancing the Search Operation Using DBpedia Ontology: Named Entity Disambiguation for Twitter Streams

N. Senthil Kumar (SITE, VIT University, Vellore, India) and Dinakaran Muruganantham (SITE, VIT University, Vellore, India)
DOI: 10.4018/IJITWE.2016040104
OnDemand PDF Download:
$30.00
List Price: $37.50

Abstract

The web and social web is holding the huge amount of unstructured data and makes the searching processing more cumbersome. The principal task here is to migrate the unstructured data into the structured data through the appropriate utilization of named entity detections. The goal of the paper is to automatically build and store the deep knowledge base of important facts and construct the comprehensive details about the facts such as its related named entities, its semantic classes of the entities and its mutual relationship with its temporal context can be thoroughly analyzed and probed. In this paper, the authors have given and proposed the model to identify all the major interpretations of the named entities and effectively link them to the appropriate mentions of the knowledge base (DBpedia). They finally evaluate the approaches that uniquely identify the DBpedia URIs of the selected entities and eliminate the other candidate mentions of the entities based on the authority rankings of those candidate mentions.
Article Preview

According to authors in (Leon Derczynski, et al, 2014), they have identified that whenever the system deal with different types of named entities, the primary task is to recognize the entities out the document collections and then classify the entities into their respective category of domains. Besides, it has to find the suitable relationship exists between the entities. Some of the Named Entity Taggers have been employed to find out the entities in different types of documents and categorized it. Further to that, in order to find the category of the entity, the favorable approach described in (R. Bunescu et al, 2006), is that they have generalized the entity types as locations, persons, organizations, timestamp, etc. While doing so, the majority of the entity types fall under the category afore mentioned. It has facilitated the process of fixing the appropriate domains to the entities and linking the entities to the correct level of meaning.

In the paper (Valentina Presutti, et al, 2014), the author has identified the candidate entities from the collection of documents, and used Wikipedia has the useful resource for identifying the potential candidate entities out of the documents. According to (Andrea Varga, et al, 2014), Once, the named entities have been extracted, ranking the entities is the crucial task which requires high governing over the entity sets. Therefore, the author has done the entity level ranking using INEX and TREC. Several approaches had been attempted thereafter in ranking the named entities but failed to fulfill the major changes to be incorporated in this methodology. Besides, the lexical similarity measures have been taken to categorize the entities to the targeted group when the entities are too ambiguous. The relevance propagation was used to filter out the entities that do not link to the set of allowed categories.

Complete Article List

Search this Journal:
Reset
Open Access Articles: Forthcoming
Volume 12: 4 Issues (2017)
Volume 11: 4 Issues (2016)
Volume 10: 4 Issues (2015)
Volume 9: 4 Issues (2014)
Volume 8: 4 Issues (2013)
Volume 7: 4 Issues (2012)
Volume 6: 4 Issues (2011)
Volume 5: 4 Issues (2010)
Volume 4: 4 Issues (2009)
Volume 3: 4 Issues (2008)
Volume 2: 4 Issues (2007)
Volume 1: 4 Issues (2006)
View Complete Journal Contents Listing