Sentiment Analysis using Legion Kernel Convolutional Neural Network with LSTM
Sukanya Ledalla1, Tummala Sita Mahalakshmi2
1Sukanya Ledalla, Assistant Professor, Department of Information Technology, Gokaraju Rangaraju Institute of Engineering and Technology University, Hyderabad (Telangana), India.
2Tummala Sita Mahalakshmi, Professor, Department of Computer Science and Engineering, GITAM Institute of Technology University, Visakhapatnam, (Andhra Pradesh), India
Manuscript received on 05 February 2019 | Revised Manuscript received on 13 February 2019 | Manuscript published on 28 February 2019 | PP: 226-229 | Volume-8 Issue-4, February 2019 | Retrieval Number: D2712028419/19©BEIESP
Open Access | Ethics and Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: Social media is growing as a communication medium where people can express their feelings online and opinions on a variety of topics in ways they rarely do in person. Detecting sentiments in texts have gained a considerable amount of attention in the last few years. Thus, the terms sentiment analysis have taken their own path to become essential elements of computational linguistics and text analytics. These terms are designed to detect peoples’ opinions that consist of subjective expressions across a variety of products or political decisions. In recent years, in India, opinions are expressed using multi-lingual words. This has become a new challenge in the area of sentiment analysis. Machine learning techniques, such as neural networks, have proven success in this task; however, there is room to advance to higher-accuracy networks. In this paper, a novel sentiment analysis system is developed which uses Legion Kernel Convolutional Neural Network with Long Short-Term Memory (LSTM). In this investigation U. S. English, Hindi dialects and datasets like twitter sentiment corpus, transliteration pairs, English word- frequency list, Hindi word-frequency list and various public opinion datasets are used. The proposed network can achieve the highest known accuracy of 92.25%. Thus the proposed network’s success can be extended to other fields also.
Keyword: Convolutional Neural Network; Long Short-Term Memory, Sentiment Analysis, Subjective Expressions, Multi-Lingual Sentence, F-Score.
Scope of the Article: Neural Information Processing