IMPLEMENTASI CHATBOT DENGAN PENDEKATAN NATURAL LANGUAGE PROCESSING UNTUK WEBSITE MANGCODING

The advancement of information technology has influenced how organizations deliver services to users. Customer service involves activities aimed at ensuring satisfaction through the assistance provided by an individual to resolve issues and meet customer needs. In certain situations, customer service may not be available outside working hours or for repetitive general inquiries. In this context, chatbots have emerged as a promising solution to enhance service quality. This research aims to implement a chatbot using Natural Language Processing (NLP) and Naive Bayes classification methods to improve the performance of Mangcoding's web- based services. The system development methodology in this study includes analyzing user requirements, designing the chatbot architecture, developing the NLP model, and integrating it with existing web service platforms. The use of NLP methods using NLTK and Naïve Bayes using Scikit learn can enhances the chatbot's ability to understand natural language and provide relevant responses according to user requests. The research employs a Mixed Methods approach to evaluate the chatbot's performance in enhancing web-based services. The chatbot is trained with data containing 13 topics or intents. Testing the chatbot model using NLP and Naive Bayes classification with a confusion matrix yielded 100% accuracy, precision, recall, and F1 score. Additionally, in terms of response time, the chatbot achieved a speed of only 0.076 seconds. The findings contribute to improving the efficiency of web-based services, supported by positive responses from 30 respondents in a questionnaire and through the implementation of the chatbot using NLP and Naive Bayes classification. This research provides guidance for organizations on leveraging chatbot technology to enhance user interaction in web-based service contexts.

Keywords: Customer service, Chatbot, Natural Language Processing, Web- Based Services, Performance, System Development, Naive Bayes