Automatic Discovery of Lexical Patterns using Pattern Extraction Algorithm to Identify Personal Name Aliases with Entities

Abstract

The personal name aliases are extremely significant in information retrieval to retrieve complete information about a personal name from the web, as some of the web pages of the person may also be referred by his or her alias name / nick name / real name. There is a rapid growth in people searching where the personal name aliases are concerned. We proposed a pattern generator which includes automatic: lexical pattern extraction algorithm and attribute extraction algorithm. We exploit three data set of known Personal names (consisting of alias name, real name, and nick name), Profession and location names of a person as training semi-structured data set to efficiently extract lexical patterns. The extracted patterns are ranked according to F-Score. It conveys information related to alias names from contingency table returned by web search engine. The extracted lexical patterns (profession pattern and location name pattern) are often used to optimize candidate personal name aliases with attributes of a person availed in the contingency table, the non-frequent items are discarded from the contingency table. Next, we ranking the candidate alias in contingency table, Graph mining ranking algorithm with various similarity measures are used then to measure the strength of association between a name and a candidate alias, co-occurrence statistics are computed.