# AI Based Chatbot: An Approach of Utilizing On Customer Service Assistance

**Authors:** Rejwan Bin Sulaiman  
**arXiv:** [2207.10573](https://arxiv.org/abs/2207.10573)  
**Date:** 2022-06-18  
**Summary page:** https://www.summarizepaper.com/en/paper/2207.10573v1/

## Key points

- Providing the best customer experience is a primary concern for online firms
- Machine learning has revolutionized the approach to client service
- Implementation of chatbot systems is a solution that has gained acceptance
- Objective: Develop a chatbot system for AK Traders Ltd. that can handle complex questions and provide logical output answers using NLP, machine learning, and VSM
- In-depth analysis of various machine learning techniques to determine the best solution
- Deliver high-quality results based on user input
- Special thanks to Dr. Paul Sant for guidance throughout the project and support from AK Traders London LTD team members and friends and family
- Report structure includes introduction, research techniques, achievements, data collection methods, data preprocessing techniques, overview of machine learning algorithms used, implementation of chatbot system using Python programming language and libraries such as NLTK scikit-learn TensorFlow, evaluation of performance by comparing with existing chatbot systems and analyzing user feedbacks, conclusion summarizing achievements, limitations, future work recommendations

## Summary

Providing the best customer experience is a primary concern for online firms, and the advancement of machine learning has revolutionized their approach to client service. One solution that has gained acceptance is the implementation of chatbot systems, which provide instant and satisfactory answers to user inquiries. This thesis paper focuses on developing a chatbot system for AK Traders Ltd., a use case company. The objective is to develop a chatbot solution that can handle complex questions and provide logical output answers using natural language processing (NLP), machine learning, and the vector space model (VSM). The project involves in-depth analysis of various machine learning techniques to determine the best solution for the company. The outcome of this project aims to deliver high-quality results based on user input. Special thanks are given to Dr. Paul Sant for his guidance throughout the project as well as support from AK Traders London LTD team members and friends and family. The report structure includes an introduction discussing the definition, history, motivation, scope, aims, and objectives of the project. Chapter 2 covers research techniques available for chatbot systems and methods capable of processing complex questions with definite answers. The achievements of this project include designing, developing, and implementing a user-friendly chatbot system that improves over time based on user experience evaluation. Additionally, extensive research was conducted on natural language processing techniques and other tools necessary for implementing the chatbot system. In-depth knowledge about different machine learning techniques and algorithms was also gained. Chapter 3 focuses on data collection methods from various sources such as websites, forums, social media platforms etc., while Chapter 4 discusses data preprocessing techniques including cleaning, normalization tokenization stemming/lemmatization stop-word removal etc. Chapter 5 presents an overview of different machine learning algorithms used in this project such as decision trees (DT), random forests (RF), support vector machines (SVM), neural networks etc. Chapter 6 discusses the implementation of the chatbot system using Python programming language and various libraries such as NLTK scikit-learn TensorFlow etc., while Chapter 7 presents an evaluation of its performance by comparing it with existing chatbot systems and analyzing user feedbacks . Finally , Chapter 8 concludes by summarizing all achievements , limitations , future work recommendations . Overall , this project aims to develop a high-quality chatbot system that can effectively interact with usersand provide satisfactory answers to their inquiries .