Artificial Neural Network Utilization for Analyzing Sentiment Polarity in Electronics Product Reviews

Intelligent systems currently have been proven to provide more benefits on various aspects of human life. One of them is sentiment analysis (SA) approach. SA is a mathematical approach that allows machines to analyze the opinion polarity of the statements or documents. Generally, SA is utilized to observe the tendency of public opinion on an issue. SA can also be used on e-commerce to analyze the trend of customer statements toward a product based on the reviews given by them. Thus, SA will help e-commerce business owners to know the level of acceptance toward offered products. In this paper, we try to evaluate the artificial neural network (ANN) algorithm in conducting a SA of electronic products reviews. In this study, the ANN was designed using 1 input layer, 1 hidden layer consisting of 10 neurons, and 1 output layer consisting of 2 neurons. Our experimental results showed that the ANN had a fairly high accuracy and precision while conducting SA toward electronic products reviews that have been carried out, i.e. 70.80% and 71.07% respectively. Hence, ANN is very possible to be applied to intelligent systems that are tasked to assist e-commerce business owners in conducting SA toward feedback provided by the customers.

[1]  Qian Liu,et al.  Automated Rule Selection for Aspect Extraction in Opinion Mining , 2015, IJCAI.

[2]  Bing Liu,et al.  Sentiment Analysis and Opinion Mining , 2012, Synthesis Lectures on Human Language Technologies.

[3]  Justin Zhijun Zhan,et al.  Sentiment analysis using product review data , 2015, Journal of Big Data.

[4]  Antonio Fernández-Caballero,et al.  Emotion Detection and Regulation from Personal Assistant Robot in Smart Environment , 2018, Personal Assistants.

[5]  Walaa Medhat,et al.  Sentiment analysis algorithms and applications: A survey , 2014 .

[6]  Prarthana A. Deshkar,et al.  Review analyzer analysis of product reviews on WEKA classifiers , 2015, 2015 International Conference on Innovations in Information, Embedded and Communication Systems (ICIIECS).

[7]  H Schmidli,et al.  Advantages of Artificial Neural Networks (ANNs) as alternative modelling technique for data sets showing non-linear relationships using data from a galenical study on a solid dosage form. , 1998, European journal of pharmaceutical sciences : official journal of the European Federation for Pharmaceutical Sciences.

[8]  Erik Brynjolfsson,et al.  What it can — and cannot — do for your organization , 2017 .

[9]  Salwani Abdullah,et al.  Approaches to Cross-Domain Sentiment Analysis: A Systematic Literature Review , 2017, IEEE Access.

[10]  Tom Fawcett,et al.  An introduction to ROC analysis , 2006, Pattern Recognit. Lett..

[11]  K. Indhuja,et al.  Fuzzy logic based sentiment analysis of product review documents , 2014, 2014 First International Conference on Computational Systems and Communications (ICCSC).

[12]  Anol Bhattacherjee,et al.  Smart Systems, Smarter Living: An Empirical Study of the Building Automation System in Organizations , 2015 .

[13]  D. V. Nagarjuna Devi,et al.  A Feature Based Approach for Sentiment Analysis by Using Support Vector Machine , 2016, 2016 IEEE 6th International Conference on Advanced Computing (IACC).