A Prototype for a Standard Arabic Sentiment Analysis Corpus

Mohammed Al-Kabi1, Mahmoud Al-Ayyoub2, Izzat Alsmadi3, and Heider Wahsheh4

1Computer Science Department, Zarqa University, Jordan

2Computer Science Department, Jordan University of Science and Technology, Jordan

3Computer Science Department, University of New Haven, USA

4Computer Science Department, King Khaled University, Saudi Arabia

Abstract: The researchers in the field of Arabic Sentiment Analysis (SA) need a relatively big standard corpus to conduct their studies. There are a number of existing datasets; however, they suffer from certain limitations such as the small number of reviews or topics they contain, the restriction to Modern Standard Arabic (MSA), etc., Moreover, most of them are in-house datasets that are not publicly available. Therefore, this study aims to establish a flexible and relatively big standard Arabic SA corpus that can be considered as a foundation to build larger Arabic corpora. In addition to MSA, this corpus contains reviews written in the five main Arabic dialects (Egyptian, Levantine, Arabian Peninsula, Mesopotamian, and Maghrebi group). Furthermore, this corpus has other five types of reviews (English, mixed MSA English, French, mixed MSA and Emoticons, and mixed Egyptian and Emoticons). This corpus is released for free to be used by researchers in this field, where it is characterized by its flexibility in allowing the users to add, remove, and revise its contents. The total number of topics and reviews of this initial version are 250 and 1,442, respectively. The collected topics are distributed equally among five domains (classes): Economy, Food-Life style, Religion, Sport, and Technology, where each domain has 50 topics. This corpus is built manually to ensure the highest quality to the researchers in this field.

Keywords: SA, opinion mining, making of Arabic corpus, arabic reference corpus, maktoob yahoo!.

Received September 17, 2015; accepted October 18, 2015