Using Twitter Sentiment Analysis for Sustainable Improvement of Business Intelligence in Nigerian Small and Medium-Scale Enterprises

Abstract

For most Small and Medium-scale Enterprises (SMEs) and startups in Nigeria, conducting market research and evaluation is solely based on customers’ verbal feedback and demands, which is often flawed and highly limited. Social media (Twitter), as a business intelligence tool, can help SMEs gain insight into customers’ perspectives of their products and services through Sentiment Analysis. This research presents a Twitter Sentiment Analysis for business intelligence using three machine learning algorithms; Bernoulli Naive Bayes (BNB), Linear Support Vector Classification (SVC), and Logistic Regression (LR) algorithm, to get the text polarity of tweets (Negative/positive). The dataset used for the research was gotten from Kaggle; the tweets were processed and analyzed using Python programming libraries on Kaggle’s Jupyter Notebook cloud environment. Our result showed that Logistic Regression achieved better performance with an accuracy of 83% based on Precision-Recall evaluation metrics.