Text Comparison Based on Semantic Similarity

Abstract

Main objective of the research is to develop a method for computing text similarity incorporating both statistical techniques and semantics. The method uses algorithms, libraries, and semantic models to analyze the word usage patterns and frequency of the texts, and can model human common sense knowledge through the use of multiple algorithms. One strength of the method is its adaptability to different domains, because of the incorporation of similar semantics. Users can choose from a range of methods within the implementation, giving them flexibility in their text analysis. The research is about developing a promising method for text similarity analysis that combines the strengths of statistical and semantic approaches. The stated technique is demonstrated to get the mapping of Course Objective sentences and Programme Outcome sentences based on similar semantics as required by National Board of Accreditation.