A Comparative Study of Root -Based and Stem -Based Approaches for Measuring the Similarity Between Arabic Words for Arabic Text Mining Applications

摘要

Representation of semantic information contained in the words is needed forany Arabic Text Mining applications. More precisely, the purpose is to bettertake into account the semantic dependencies between words expressed by theco-occurrence frequencies of these words. There have been many proposals tocompute similarities between words based on their distributions in contexts. Inthis paper, we compare and contrast the effect of two preprocessing techniquesapplied to Arabic corpus: Rootbased (Stemming), and Stem-based (Light Stemming)approaches for measuring the similarity between Arabic words with the wellknown abstractive model -Latent Semantic Analysis (LSA)- with a wide variety ofdistance functions and similarity measures, such as the Euclidean Distance,Cosine Similarity, Jaccard Coefficient, and the Pearson CorrelationCoefficient. The obtained results show that, on the one hand, the variety ofthe corpus produces more accurate results; on the other hand, the Stem-basedapproach outperformed the Root-based one because this latter affects the wordsmeanings.