Sentiment Analysis for Palestinian Dialect Arabic Language
No Thumbnail Available
Date
2017
Authors
Asaad, Murad
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Sentiment Analysis is the process of determining the sentiment of a text written in natural language. A number of lexicons are available for English language. In this paper, we present a sentiment lexicon for Arabic language written Palestinian dialect. In addition, we present the process of collecting and annotating data from Facebook social media Palestinian public pages. Moreover, we present results after using machine-learning algorithms to enhance the classification prediction of Positive and Negative classes. The experiments were conducted over testing data taken from the data used to build the lexicon and we were able to achieve over 90% F-score and about 80% F-score for unseen data. is the process of determining the sentiment of a text written in natural language. A number of lexicons are available for English language. In this project, we present a sentiment lexicon for Arabic language written Palestinian dialect with real time classification using Apache spark and Elasticsearch. In addition, we present the process of collecting and annotating data from Facebook social media Palestinian public pages. Moreover, we present results after using machine-learning algorithms to enhance the classification prediction of Positive and Negative classes. The experiments were conducted over testing data taken from the data used to build the lexicon and we were able to achieve over 90% F-score and about 80% F-score for unseen data.
Description
Keywords
Sentiment Analysis, Sentiment Lexicon, Data Analysis, Text classification, machine learning, Spark, Elasticsearch