Framework for pattern generation from discriminating datasets
by A. Muthusamy; A. Subramani
International Journal of Collaborative Intelligence (IJCI), Vol. 1, No. 2, 2015

Abstract: Searching for the exact name of a person on the web is a challenging task when a single name is shared by many people. The proposed method is based on alias detection to extract the lexical pattern, attributes of a person and candidate alias ranking. Initially we construct training semi-structured datasets which consist of alias name or nickname or real name; the profession and location names of a person are framed with the help of social media networks such as Wikipedia. Secondly, pattern generator includes pattern extraction algorithm and attributes extraction algorithm. Pattern extraction is used to generate lexical patterns manually with the use of dataset, and then generated patterns are passed to the Google search engine. According to the evaluation scheme the extracted lexical patterns are ranked. For ranking the candidate alias of a person, a graph mining ranking algorithm with various similarity measures is used. This paper presents the overview of the people search engine, definition, proposed model, training semi-structured dataset and evaluation scheme of manually created lexical patterns.

Online publication date: Mon, 17-Aug-2015

The full text of this article is only available to individual subscribers or to users at subscribing institutions.

 
Existing subscribers:
Go to Inderscience Online Journals to access the Full Text of this article.

Pay per view:
If you are not a subscriber and you just want to read the full contents of this article, buy online access here.

Complimentary Subscribers, Editors or Members of the Editorial Board of the International Journal of Collaborative Intelligence (IJCI):
Login with your Inderscience username and password:

Username:        Password:          Forgotten your password?

Want to subscribe?
A subscription gives you complete access to all articles in the current issue, as well as to all articles in the previous three years (where applicable). See our Orders page to subscribe.

If you still need assistance, please email subs@inderscience.com