Sentiment Analysis in Customer Experience in Philippine Courier Delivery Services using VADER Algorithm Thru Chatbot Interviews

Abstract

The use of sentiment analysis of the customers' insight on the product delivery services would be an excellent opportunity to evaluate the customers' emotions in the courier delivery services. The result can link with the review to substitute for the product's performance or customer satisfaction. The primary steps in the study are data collection, data pre-processing, and sentiment analysis. The emotional data were analyzed using the VADER algorithm. A chatbot was used as an intermediary tool to collect emotional datasets. The compound score was calculated by adding the valence scores of each word in the lexicon, then adjusting them according to the guidelines and normalizing them. The negative, neutral, and positive scores represent the customer's sentiment. The test results achieved 93.33% sentiment accuracy based on the computed sentiment polarities and compound scores. The output evaluates that the system is effective based on the outcome and can be considered an innovative approach in analyzing the customers' perception of the product and services available in the market.