文章基本信息

标题：AN AUTOMATIC LEXICON WITH EXCEPTIONAL-NEGATION ALGORITHM FOR ARABIC SENTIMENTS USING SUPERVISED CLASSIFICATION
本地全文：下载
作者：AYMAN MOHAMED MOSTAFA
期刊名称：Journal of Theoretical and Applied Information Technology
印刷版ISSN：1992-8645
电子版ISSN：1817-3195
出版年度：2017
卷号：95
期号：15
页码：3662
出版社：Journal of Theoretical and Applied
摘要：Sentiment analysis is a kind of natural language processing that determines the feelings of people in a piece of text they are positive, negative, or neutral. Analysis of Arabic sentiments is considered a complex task due to the large linguistic and negation terms in Arabic language. Most recent researches are based on detecting the polarity term after the negation particle immediately. This can reduce the accuracy and performance of the analysis because many sentiments especially written in slang Arabic do not depend on having a negation particle before the polarity term. The aim of this paper is to develop a hybrid sentiment classification based on automatic lexicon algorithm and machine learning approach. The automatic lexicon is developed with a negation algorithm for both modern standard Arabic and colloquial Arabic. This algorithm detects the negation particle and traces all polarity terms even if they do not come after the negation particle. An exceptional negation is embedded into the negation algorithm which is based on the Arabic exceptional pattern in reversing the polarity term after the negation process. The experimental results are conducted using supervised machine learning methods such as SVM, KNN and NB that achieve high results in accuracy, precision, recall, and F-measure which are compared with the experimental results in three recent research papers.
关键词：Sentiment Analysis; Automatic Lexicon; Machine Learning; and Exceptional Negation