Semantic Text Similarity

Abstract

Semantics has been, in recent years, an open area of study in the field of information retrieval. Considering supervised classification of text, which is the main focus of this work, semantics is involved in different stages of processing of text: during the indexing stage, during the training phase, and via class prediction step. New foundational similarity measurements from text to text may replace classical similarity measures commonly used with other decision-making classification methods. Here is a brand-new method to test semantic resemblance using the Corpus-based Approach which is sponsored for a text as a brand-new feature that summarizes semantic resemblance between concepts representing the pair-to-pair compared text documents.I.