Adam Deep Learning With SOM for Human Sentiment Classification

• Published in: International journal of ambient computing and intelligence Vol. 10; no. 3; pp. 92 - 116
• Main Authors: Dey, Nilanjan, Ali, Md Nawab Yousuf, Sarowar, Md Golam, Rahman, Md Lizur, Chaki, Jyotismita, Tavares, João Manuel R.S
• Format: Journal Article
• Language: English
• Published: Hershey IGI Global 01.07.2019
• Subjects: Big Data; Classification; Comparative analysis; Comparative studies; Computational linguistics; Deep learning; Dictionaries; Feasibility; Language processing; Machine learning; Natural language interfaces; Natural language processing; Neural networks; Principal components analysis; Product reviews; Public opinion; Self organizing maps; Sentiment analysis; Social networks; Support vector machines; Task complexity; Unsupervised learning; India
• ISSN: 1941-6237; 1941-6245
• DOI: 10.4018/IJACI.2019070106

Nowadays, with the improvement in communication through social network services, a massive amount of data is being generated from user's perceptions, emotions, posts, comments, reactions, etc., and extracting significant information from those massive data, like sentiment, has become one of the complex and convoluted tasks. On other hand, traditional Natural Language Processing (NLP) approaches are less feasible to be applied and therefore, this research work proposes an approach by integrating unsupervised machine learning (Self-Organizing Map), dimensionality reduction (Principal Component Analysis) and computational classification (Adam Deep Learning) to overcome the problem. Moreover, for further clarification, a comparative study between various well known approaches and the proposed approach was conducted. The proposed approach was also used in different sizes of social network data sets to verify its superior efficient and feasibility, mainly in the case of Big Data. Overall, the experiments and their analysis suggest that the proposed approach is very promissing.