Paraphrase Identification Between Two Sentence Using Support Vector Machine

Abstract

Paraphrasing sentence is to express a sentence using another form of language without changing the meaning of the previous sentence. In this study, a system of identification of meanings of sentences has been built using Support Vector Machine (SVM) with Sequential Minimal Optimization (SMO) algorithm, the features used are Euclidean distance, cosine similarity, sentence length 1, sentence length 2, slices between sentences. The test results showed better results on SVM configuration using attribute filters achieving the best results 94.4% in training using the train set test and 61.9% in SVM configuration without using attribute filters.