DESIGN AND IMPLEMENTATION OF AN ONLINE CHATBOT FOR ADMISSION ENQUIRY

CHAPTER ONE

INTRODUCTION

Background to the study: The usage of chatbots in various business sectors has been on the rise in recent years, and higher educations like the university; is no exception. In higher educational context, applicants sends their queries through the university’s email or the inquiry section on the website which most times are slowly responded to or left unanswered, while chatbots can provide an accurately and quick responses to common admission enquiries, reducing the workload of the admissions team and improving the overall experience for applicants. However, to implement this chatbot which can be complex, requires the understanding of natural language processing (NLP) and machine learning (ML) technologies, as well as thorough understanding of the user’s journey and the various touchpoints that prospective students have with the university during the enquiry process. Additionally making the interaction between the bot and the human more humanly with the use of conversational AI. A chatbot system simulate a discussion (or a chat) with a user in natural language using conversational artificial intelligence (AI) technology via messaging applications, websites, mobile apps or the telephone (Selig, 2022). A chatbot is artificially constructed software that uses natural language as input and output to talk to humans. Chatbots can act as a personal assistant on mobile devices to provide users with personalized information, enable real-time social interaction media, and can even be used in health consultations (M. Yamada, 2016).

According to (Court Bishop, 2022) Customer service teams that handles upto 20,000 support requests on a monthly basis could save more than 240 hours per month by just using chatbots. Chatbots are used increasingly in instant messaging and they are implemented in people's regular lives, shopping experiences, and education courses (Ferrell, 2020). In general bots are just machines that is intelligent enough to understand human request and then formulate the request in a way that is understandable by other software systems to request the data you need (Sumit, 2019). This study aim to address the gap of prompt response between the university and prospective students and evaluating its effectiveness in improving the overall experience for applicants and streamlining the enquiry process for the admissions.

1.2 Statement of problem

For every new academic session, the university receives many applications which are processed into enrolment, applicants may need to ask questions ranging from courses, the university’s environment, and fees to accommodation and learning systems. With so much queries from the students, putting in much workload on the admission teams, most queries may be left unanswered. There is a need for a system which can provide accurate and very quick responses to enquiries, and guide applicants on what to do, by handling multiple enquiries at once.

1.3 Aim and objectives of study

The aim of this study is to develop a system that provides quick and accurate response to queries from applicants and reducing the workloads of the admissions team. The specific objectives are to:

1. collect data relevant to the student applications and it features

2. formulate the model using the features in (i)

3. design the model

4. implement the system.

5. test the system.

1.4 Methodology of the study

To accomplish the aforementioned objectives fully, the waterfall model is used. The waterfall model contains series of processes, that are used duing development. The stage usually will require:

a. Gathering of requirement for the proposed system which will involve understanding of. the user journey and the various touchpoints that prospective students have with the university during the enquiry process, like answers to the queries, logs, feedback, keywords.

1. A file called intents will be created which will contain every features extracted in (a) . These features will be grouped according to how the bot will be trained. That is; Interactive FAQ, question answering, dialogue planning etc

2. Design of the proposed system will be specified using UML diagram such as Use case, Sequence diagram, Activity diagram.

3. The implementation will be done using HTML, CSS, jQuery and Ajax api, JavaScript as the frontend and Python Language and Natural Language Processing (NLP) as the backend with PostgreSQL as the database.

4. The system will be tested by users via a user assessment test questionnaire.

1.5 Scope and Limitation of the study

This scope of this study is to develop a web-based chatbot that will be used to manage the common enquiries of the applicants, increase efficiency and reduce response time during multiple enquiries, availability 24/7 etc. Due to time availability for this study, the chatbot is an Information-retrieval bot. Therefore, the chatbot cannot self-learn or generate it own newly response.

1.6 Significance of the study

This study will reduce the workload of the admissions team by providing quick and accurate responses to common enquiries, it will handle multiple enquiries at once and be available 24/7. It will also reduce the cost of operation and increase the reach of the University to many student who are looking for information.

1.7 Defination of terms

1. Applicants: An applicant is one who formally applies for or requests something, especially a job or to study at a university or college.

2. Artificial Intelligence: It is the effort to automate intellectual tasks typically performed by humans. It can be complex or simple. A computer-controlled robot can perform duties normally related to intelligent beings.

3. Bootstrap: Bootstrap is a CSS Framework used in developing responsive and mobile-first websites.

4. Chatbot: A chatbot is a compound word of a chat and a bot. It is an application that simulates and strategies human verbal exchange (both written and spoken), permitting them to engage with virtual gadgets as if they had been speaking with an actual individual.

5. Conversational AI: It is a combination of machine learning and natural language processing that allows people to have huma-like interactions with computer.