Statistical learning approaches to sentiment analysis in the Nigerian banking context

Banking is an essential component of our day-to-day activity, and the sector contributes to the development of every Nation’s economy. To ensure bank stability, the banks have gone through several deregulation and reformation processes. Unfortunately, these processes and technological advancement have led to increased competitiveness, saturated market, and low profitability. Thus, the need for banks to build customer centric service to gain profitability and stability is vital. This opens the need to investigate customers attitude towards banking. With the growing usage of social network sites (SNS), the user generated content (UGC) has opened opportunities for banks to mine customers opinion. This can help the banks to generate insight into their product and services, create marketing strategies and manage their reputation. To the customers, this serves as source to information that can help in supporting decision making on product or service purchase. In this study, sentiment analysis (SA) techniques were employed to investigate customers attitude towards banking using Twitter data. However, the unstructured nature of the data and word ambiguity made sentiment analysis complicated. In the context of this study, SA is more difficult because it involves the natural language processing of Pidgin English and English words in the bank domain. Unfortunately, there are limited or no resources for this purpose. This study is of two main tasks namely, sentiment classification and aspect extraction for sentiment analysis. For the sentiment classification task, this study utilised both lexical based approach and the machine learning approach. The lexical based approach relies on opinion words, it is easy to understand the sentiment classification result and ways to improve. Machine learning algorithms are black box models and often are not interpretable by a human. However, they produce models with good performance. The study proposes SentiLeye, a novel lexicon algorithm and compared with existing lexicons. Results showed SentiLeye outperformed others due to domain terms, opinionated-objective words, negation, and language which was put into consideration during the development. Alternatively, the machine learning models were compared. The performance result of the classification models validated Support Vector Machine (SVM) with accuracy of 82% as the most performed classification model in the banking context. The second task involved the use of statistical topic modelling techniques for aspect extraction. The topic models were compared, and Latent Dirichlet Allocation (LDA) showed the best performance in terms of topic coherence and interpretable terms. Thus, this study proposes a Topic-Sentiment Banking System (TSBS) framework which was used to demonstrate the aspect of banking which customers were happy or unhappy with. Our result showed the significance of customer service experience, transaction problem, and bank charges. Thus, recommends the banks to pay attention to these topics as our findings showed significant proportion of customers are unhappy with these aspects of banking.