markbosire/Kenya-Airways-customer-feedback-analysis

A notebook that will do a customer feedback analysis on kenya airways comments # Kenya Airways Custommer Feedback Analysis This repository contains a Jupyter Notebook (`Kenya Airways Reviews.ipynb`) that performs sentiment analysis and emotions analysis on reviews of Kenya Airways. The data used for analysis is scraped from Airline Quality. ## Notebook Content ### Data Cleaning and Preprocessing - The notebook begins with importing the necessary libraries and the dataset. - Text preprocessing functions are defined to clean the text data, including removing useless text, converting text to lowercase, removing punctuation, tokenization, and handling stop words. ### Emotion Analysis - Emotion dictionary function is created to map emotions from a file into a dictionary. - Sentiment analysis using Vader Sentiment Analyzer is implemented. ### Plot Findings - Visualizations are created to summarize the sentiment distribution, frequent emotions, frequent categories, and aspect-based sentiment analysis. - The visualizations include pie charts, bar plots, and word clouds to represent the findings effectively. ### Aspect-based Sentiment Analysis using BERT - Pre-trained BERT model is utilized for aspect-based sentiment analysis. - Aspects like flight, service, seat, food, and crew are extracted from the text, and sentiment analysis is performed for each aspect. ## Usage - Ensure you have the necessary dependencies installed as specified in the notebook. - Run each cell sequentially to execute the analysis. - The notebook provides detailed comments and explanations for better understanding. ## Libraries Used - pandas - NLTK - Matplotlib - WordCloud - transformers - spacy ## Note - Some parts of the notebook require downloading NLTK data and emotion dictionary files. - The notebook provides comprehensive analysis and visualization of Kenya Airways reviews, aiding in understanding customer sentiments and emotions associated with different aspects of the airline service.