Analysis of Anchor Text Based on Pattern Growth Graph Algorithm for Name Alias Detection System

Identifying the correct alias for person's name playing a crucial role in the field of information retrieval, sentiment analysis, and person name disambiguation as well as in biomedical fields. Traditional system provides the solution on solving lexical ambiguity, but it lagged on the problem of referential ambiguity. Through this paper we emphasis on referential ambiguity to extract correct alias for a given name. Given a person name and/or with context data such as location, organization retrieves top-K snippets from a web search engine. With the help of Lexical-pattern extract candidate aliases. As to find correct alias from a list of aliases we used anchor text analysis based on link and forming graph with link called as in-link and out-link. Anchor text analysis used co-train algorithm for preprocessing and after that prepared a set of anchor text word. For rank a node from graph we integrate various similarity measures such as dice, Jaccard coefficient for word relation along with degree distribution and clustering coefficient. There by our method providing more promising result in terms to improve the precision and minimize the recall that than the previous baseline method.