A Comparison of Sentiment Analysis Algorithms Using Twitter Feedback on a Local Bank: Developing Marketing Strategies for Philippine Banks

Authors: Alexander Manuel Dean T. Ferrazzini, Leo Neil P. Flores, Justin Jay L. Joaquin, Cherry B. Lisondra

DOI: https://doi.org/10.37082/IJIRMPS.IPMESS-24.7

Short DOI: https://doi.org/mgfc

Full-text Research PDF File:   View   |   Download

Abstract: A breach in BDO's online security has escalated into clients losing their money, and BDO is faced with the need to provide assurance that strict cyber security protocols are being followed and that the risk of outsider hacking is negligible. Therefore, sentiment analysis is an appealing tool to facilitate the collection of feedback to improve banking processes. To perform the analysis, several Python Libraries and the Naïve-Bayes Algorithm were used. For data mining, Apify was used via its Twitter Scraper feature to gather data for the dataset used to train the algorithm. The individual accuracy of the program created was then compared to existing studies that used the Semantic Approach and J48 Algorithm to determine which algorithm is the most accurate and superior. Compared to J48 and Semantic Approach, J48 is the most accurate with Semantic Approach being the least accurate. The accuracy of the Naive Bayes algorithm is in the middle at 86.82%. While not as accurate as J48, this is compensated by its faster processing time and faster learning capabilities. From the results of the sentiment analysis, BDO must improve in various facets of their service such as cybersecurity, customer service, and privacy. The website and app must also be optimized for better online banking. For future research, it is recommended that the dataset is processed in other algorithms for a more accurate comparison. Front-end integration should also be considered for non-technology adept users and the code must also be optimized for efficient and more accurate data processing.

Keywords: Bank, Sentiment Analysis, Marketing, Feedback

Paper Id: 4.207

Published On: 2024-01-30

Published In: Special Issue - International Conference on Innovative Practices in Management, Engineering & Social Sciences (January 2024)

Share this