Sentiment Analysis

Showing results in 'Publications'. Show all posts
Siddiqui, S., A. A. Monem, and K. Shaalan, "Sentiment Analysis in Arabic", Natural Language to Information Systems: 21st International Conference on Applications of Natural Language to Information Systems (NLDB 2016), Berlin , Springer, 2016. Abstractsentiment_analysis_in_arabic.pdf

The tasks that falls under the errands that takes after Natural Language Processing approaches includes Named Entity Recognition, Information Retrieval, Machine Translation, and so on. Wherein Sentiment Analysis utilizes Natural Language Processing as one of the way to locate the subjective content showing negative, positive or impartial (neutral) extremity (polarity). Due to the expanded utilization of online networking sites like Facebook, Instagram, Twitter, Sentiment Analysis has increased colossal statures. Examination of sentiments helps organizations, government and other association to extemporize their items and administration in view of the audits or remarks. This paper introduces an Innovative methodology that investigates the part of lexicalization for Arabic Sentiment examination. The system was put in place with two principles rules– “equivalent to” and “within the text” rules. The outcomes subsequently accomplished with these rules methodology gave 89.6 % accuracy when tried on baseline dataset, and 50.1 % exactness on OCA, the second dataset. A further examination shows 19.5 % in system1 increase in accuracy when compared with baseline dataset.

Atia, S., and K. Shaalan, "Increasing the Accuracy of Opinion Mining in Arabic", The International Conference on Arabic Computational Linguistics (ACLing), Cairo, Egypt, 18 April, 2015. Abstractopinionminacling2015.pdf

Opinion Mining is a raising research field of interest, with its different applications derived by market needs
to analyze product reviews or to assess the public opinion, for political reasons, during presidential campaigns. In this paper, we address an approach for improving accuracy of Opinion Mining in Arabic. In order to conduct our study we need Arabic linguistic resources for opinion mining. Investigating the available resources we found that the OCA corpus is available and sufficient to prove our approach. Experimental results showed that applying different parameters of the machine learning classifiers on the OCA corpus leads to increasing the accuracy of the Arabic Opinion Mining.