Deep Recurrent Network Based Feature Selection using Single Matrix Normalization and Eigen Vectors for Analyzing Sentiments
Rajalaxmi Hegde1, Sandeepkumar Hegde2
1Rajalaxmi Hegde, Department of CSE, NMAM Institute of Technology, Nitte, India.
2Sandeepkumar Hegde, Department of CSE, NMAM Institute of Technology, Nitte, India.
Manuscript received on 02 July 2019 | Revised Manuscript received on 06 July 2019 | Manuscript published on 30 August 2019 | PP: 804-809 | Volume-8 Issue-10, August 2019 | Retrieval Number: J89130881019/2019©BEIESP | DOI: 10.35940/ijitee.J8913.0881019
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: Sentiment analysis plays a major role in e-commerce and social media these days. Due to the increasing growth of social media, a huge number of peoples and users send their reviews through the Internet and several other sources. Analyzing this data is challenging in today’s life. In this paper new normalization based feature selection method is proposed and the topic of interest here is to select the relevant features and perform the classification of the data and find the accuracy. Stability of the data is considered as the most important challenge in analyzing the sentiments. In this paper investigating the sentiments and selecting the relevant features from the data set places a major role. The aim is to work with the vector-based feature selection and check the classification performance using recurrent networks. In this paper, text mining depends on feature retrieval methods to improve accuracy and propose a single matrix normalization method to reduce the dimensions. The proposed method performs data preprocessing or sentiment classification and features reduction to improve accuracy. The proposed method achieves better accuracy than the N-gram feature selection method. The experimental results show that the proposed method has better accuracy than other traditional feature selection approaches and that the proposed method can decrease the implementation time.
Keywords: Features, Classifier, Sentiment, Normalization, Recurrent Networks, Covariance Matrix, Cosine similarity.
Scope of the Article: Network Based Applications