SUPERVISED ALIAS NAME VALIDATION USING STATISTICAL SIMILARITY COEFFICIENTS

Abstract

Alias name is the surnames for a known name. Extracting and validating alias names is an interesting research topic in language processing and has a number of Natural language processing applications like Information extraction, Information retrieval, Sentimental analysis, Question and answering. Alias name validation involves the process of validating whether a name is alias name or not. In this work, seven statistical similarity coefficients were used as features in classifier to validate alias names. For each name-alias pair, seven statistical similarity coefficient values were calculated and used as features to train a classifier. The trained classifier is then employed to classify whether a name-alias pair is valid or not. Experiments were conducted using Indian name-alias data that has data for 15 persons containing 35 name-alias pairs. Results show that SVM classifier with Radial Basis Function Kernel outperforms all the other classifiers in terms of overall accuracy.

Keywords

• Alias name extraction
• Information Extraction
• Web Mining