Abdulraqib20/Banking-Sentiment-Analysis-in-Nigeria

The aim of this project is to perform sentiment analysis on Twitter data related to customer satisfaction with Nigerian banks, in order to gain insights into customer opinions and experiences and identify areas for improvement in the banking industry. # Banking-Sentiment-Analysis-in-Nigeria ## Description The aim of this project is to perform sentiment analysis on Twitter data related to customer satisfaction with Nigerian banks, in order to gain insights into customer opinions and experiences and identify areas for improvement in the banking industry. By collecting and analyzing tweets mentioning Nigerian banks, we aim to understand the sentiments expressed by customers, whether positive or negative, towards various banking services and experiences. Through natural language processing techniques and the application of XGBoost, Support Vector Machine Classifier models, BERT and KerasNLP , we seek to accurately classify these sentiments and uncover valuable patterns and trends. The findings from this analysis can provide valuable business intelligence to Nigerian banks, enabling them to make data-driven decisions, optimize customer service strategies, and address customer concerns more effectively. ## Key Features - Introduction: Overview of the project. - Data Collection: Collected data from Twitter by querying the Twitter API. - Data Cleaning: Clean and preprocess the collected data for analysis. - Data Preprocessing: Perform text preprocessing techniques such as removal of URLs, hashtags, mentions and special characters and tokenization, stop word removal, and stemming/lemmatization. - Feature Engineering: Additional features such as character count and word count were considered, and new features were engineered from the dataset, such as time, hour, weekday, month, and year, to provide supplementary insights into the tweet structure and length. - Sentiment Scoring: Calculating the sentiment scores and classifying sentiments into two distinct classes (positive and negative). - Analyzing keywords and phrases used in the dataset. - Exploratory Data Analysis: Conducted comprehensive EDA on the pre-processed data to gain valuable insights into sentiment distribution trends, patterns, and potential anomalies. Thi …