A Review on Sentiment Analysis Techniques and Applications

Joint Conference on Green Engineering Technology & Applied Computing 2019

工业技术（总论）;计算机科学

Ridzwan Yaakub, Mohd^1 ; Iqbal Abu Latiffi, Muhammad^1 ; Safra Zaabar, Liyana^1

Center for Artificial Intelligence Technology, Faculty of Information Science and Technology, Universiti Kebangsaan Malaysia, 43600 UKM Bangi, Selangor, Malaysia^1

关键词: Analysis techniques;    Classification tasks;    High-accuracy;    Machine learning techniques;    NAtural language processing;    Test data;   

Others  :  https://iopscience.iop.org/article/10.1088/1757-899X/551/1/012070/pdf
DOI  :  10.1088/1757-899X/551/1/012070

来源: IOP

【 摘 要 】

Nowadays, what user think is the most difficult and complicated task handled by organizations. The way to idetify the attitude of the speaker or a writer on some topics is to use sentiment analysis. The use of sentiment analysis is to identify user's opinion towards some topics whether it is positive or negative. This paper presents the techniques used by previous researchers in sentiment analysis which are Machine Learning and Natural Language Processing (NLP) in solving the classification task. The comparison among these two main approaches reveals that Machine Learning techniques can solve classification task with reasonable success and with very high accuracy compared to NLP-based techniques but it is depending on the training and test data with respect to the domain. This paper also presents the use of ontology in sentiment analysis that can help in achieving more high accuracy for the classification task.