Using chatbot to provide better customer service for the union of industry, finance and commerce workers (unicof)

Abstract

This dissertation presents the development and deployment of UNIBOT, a chatbot designed to enhance customer service within the Union of Industry, Commerce, and Finance Workers (UNICOF). This work is being done on the backdrop of rapid technological change and its influence on the corporate landscape and also the importance of customer engagement for business success. The research attempts to highlight the use of artificial intelligence and in this specific case chatbot to enhance the customer service experience. The research methodology employed a text-based rule-based approach, utilizing Python and the Natural Language Toolkit (NLTK). A structured database, derived from UNICOF's website and staff input, formed the foundation for the UNIBOT's knowledge. In addition, tokenization, pre-processing, and sentence generation techniques were implemented. The methodology also explains the architecture design and implementation steps, encompassing tokenization, greeting functions, and query responses of the UNIBOT. Results confirm the successful execution and testing of UNIBOT, evidencing its ability to effectively respond to user inquiries. Conversations with UNIBOT reflect informative and accurate responses interactions ranging from greetings to specific questions about UNICOF. However, challenges arising from the lack of specific frequently asked questions (FAQ) data and website integration are acknowledged.