Exploiting Machine Learning for Comparative Sentences Extraction

Abstract

This paper studies the problem of extracting Chinese comparative sentences from user reviews, which is a problem of text classification in the level of sentence. This paper first deals with the class skewed problem of review data, and then builds a SVM (support vector machine) model to classify comparative and non-comparative sentences into different groups on a balanced dataset. Various linguistic and statistical features are introduced to characterize a sentence. Experiments were conducted on user-generated product reviews. As a result, our experiments show significant performance, an overall F-score of 85.87%.