Optimal Selection of Python to Create

the ChatBot
A minor project report for the evolution and partial fulfillment of
the requirement for the award of the degree MCA(AI)

SUBMITTED BY

SHUBHAM KUMAR (235/PCA/024)

ANUJ KUMAR (235/PCA/004)
MRADUL KUMAR (235/PCA/016)
YUVRAJ SAINI (235/PCA/032)

UNDER THE SUPERVISION OF

MR. ATUL KUMAR

SCHOOL OF INFROMATION AND COMMUNICATION TECHNOLOGY

GAUTAM BUDDHA UNIVERSITY
GREATER NOIDA-201312, UTTAR PRADESH, INDIA
OCTUBER, 2024

[1]
 SCHOOL OF INFROMATION AND COMMUNICATION
TECHNOLOGY
GAUTAM BUDDHA UNIVERSITY, GREATER NOIDA, 201312

CANDIDATE’S DECLARATION

We hereby certify that the work embodied in this project report entitled
“Optimal Selection of Python to Create the ChatBot” is submitted for the
partial fulfillment of the requirements for the award of the Degree of
MCA(AI) submitted to the School of Information and Communication
Technology, Gautam Buddha University, Greater Noida. It is an authentic
record of our own work carried out under the supervision of Mr. Atul
Kumar , School of ICT. The matter presented in this report has not been
submitted in any other University/ Institute for the award of any other
degree or diploma. Responsibility for any plagiarism related issue stands
solely with us.

Name and roll numbers:-

Shubham kumar (235/PCA/024)

Anuj Kumar (235/PCA/004)
Mradul Kumar (235/PCA/016)
Yuvraj Saini (235/PCA/032)

Signature:

Signature of the Supervisor:

[2]
 SCHOOL OF INFROMATION AND COMMUNICATION
TECHNOLOGY
GAUTAM BUDDHA UNIVERSITY, GREATER NOIDA, 201312

CERTIFICATE

This is to certify that the work done entitled “Optimal Selection of

Python to create ChatBot ”, by “Shubham Kumar ” Roll Number
“235/PCA/024”,“Mradul Kumar” Roll Number “235/PCA/016”, “Anuj
Kumar” Roll Number “235/PCA/004”, “Yuvraj Saini” Roll Number
“235/PCA/032” Master’s in Computer Application Second year has been
carried under my supervision.

Mr. Atul Kumar

Gautam Buddha University Greater Noida, 201312
(U.P.) INDIA

[3]
 ACKNOWLEDGEMENT

We have taken efforts in this project. However, it would not have been
possible without the kind support and help of many individuals and
organizations. I would like to extend my sincere thanks to all of them.

We are highly indebted to Mr. Atul Kumar for her guidance and
constant supervision as well as for providing necessary information
regarding the project & also for her support for completing the project.

We would like to thank Dean of ICT Prof. S. K. Sharma and HOD of

IT, Dr. Neeta Singh whose support helped me to complete my work.
We would like to express our gratitude towards our friends and
family members for their kind cooperation and encouragement which
helped us in the completion of this project.

Name and Signature:

Shubham kumar (235/PCA/024)

Anuj Kumar (235/PCA/004)
Mradul Kumar (235/PCA/016)
Yuvraj Saini (235/PCA/032)

[4]
 CONTENT
CANDIDATE’S DECLARATION..........................................................................................................2
CERTIFICATE.........................................................................................................................................3
ACKNOWLEDGEMENT........................................................................................................................4
CONTENT................................................................................................................................................5
ABSTRACT..............................................................................................................................................6
1. Introduction
1.1.Purpose of the Program
1.2.Technologies Used
2. System architecture

3. Setting up the Environment

3.1.Installation of Required Libraries
3.2.Overview of the Code Components
4. Voice Input with Speech Recognition
4.1.Overview of Speech Recognition Library
4.2.Code for Capturing Voice Input
4.3.Handling Errors in Speech Recognition
5. Text Generation with OpenAI API
5.1.OpenAI API Integration
5.2.Explanation of Model Selection
5.3.Prompt Construction and Completion
6. Voice Output with pyttsx3
6.1.Introduction to pyttsx3 Library
6.2.Synthesizing Speech Output
6.3.Handling Voice Responses in the Program
7. Code Workflow
7.1.Step-by-Step Breakdown of Code Flow
7.2.Listening to User Input
7.3.Interacting with OpenAI
7.4.Outputting the Response
8. Error Handling and Robustness
8.1.Handling Speech Recognition Errors
8.2.Error Handling with OpenAI API
8.3.Improving Program Stability
9. User Interaction and Looping
9.1.Continuous Listening Loop
9.2.Ending the Program Gracefully
10. Use Cases
10.1. Applications of Voice Recognition and AI Text Generation
10.2. Practical Scenarios

11. Performance Optimizations

12. Conclusion
[5]
13. REFERENCES

[6]
 ABSTRACT

A chatbot, usually referred to as a chatterbot, attempts to have a conversation with a

person. When a question is posed, the system has the ability to detect sentences and select
the proper answer. The response principle is the matching of the user's input phrase. The
current technical project involves building a professional system for a college help desk
employing an android-based chatbot, artificial intelligence technology, and virtual
assistance (human-machine communication), then sending that natural language to a
server. Chatbot systems have become increasingly popular for automating interactions with
users and providing information in various domains, including college enquiries. In this
paper, we propose a chatbot system for college enquiry using a knowledgeable database.
The system utilizes a knowledgeable database that contains relevant information about the
college, such as courses, faculty, campus facilities, and admissions procedures. The system
employs various algorithms, including rule-based, retrieval-based, natural language
processing (NLP), and machine learning algorithms, to understand and respond to user
queries in a context-aware manner. The rule-based algorithms provide predefined rules and
patterns for handling specific intents or frequently asked questions, while the retrieval-
based algorithms search the knowledgeable database for relevant informations.

[7]
 1. INTRODUCTION
This section provides an overview of voice recognition, natural language processing (NLP), and
the integration of OpenAI’s API with speech recognition and text-to-speech technologies for
interactive systems.
This Application is for college students, staff, and parents. Easy way to interaction and time
consuming. This project is mainly targeted at colleges and the synchronization of all the sparse
and diverse information regarding regular college schedule. Generally, students face problems
in getting correct notifications at the correct time, sometimes important notices such as campus
interview, training and placement events, holidays, and special announcements. Smart Campus
tries to bridge this gap between students, teachers, and college administrators. Therefore in the
real world scenario, such as college campus, the information in the form of notices, oral
communication, can be directly communicated through the android devices and can be made
available for the students, teachers directly for their android devices and the maintenance of
application will be easier in later future because of the use of architectural MVC which
separates the major works in the development of an application such as data management,
mobile user interface display and web service which will be the controller to make sure for fast
and efficient maintenance of application. The College bot project is built using artificial
algorithms that analyses user’s queries and understand user’s message.

1.2 PROBLEM STATEMENT

At the start of each academic semester, registration opens for those wishing to join the
university in various disciplines, and telephone calls for admission and registration abound.
This leads to an increase in the loads and work for the employees of the Deanship of Admission
and Registration as a result of the constant pressure of those wishing to register and their
families by flocking to the Deanship, so the employees are not able to answer the phone calls
and social media. This often leads to many students who wish to register to be ignored. The
process of providing information and support to prospective and current students in a timely
and efficient manner is a challenge for colleges, leading to frustration and dissatisfaction among
users. To achieve this, the chatbot system must be built on a robust and comprehensive 3
knowledge database that contains all relevant information about the college and its operations.
This database should be regularly updated to ensure that the information provided by the
chatbot is accurate.

[8]
 1.3 OBJECTIVES 
 Save effort and time for both the admission and registration staff and students who wish to
enroll. 
 Provide detailed information about colleges and majors. 
 Easy access to information. 
 To minimize the time required to solve the queries. 
 To give response to the user based on queries. 
 To simplify communication between user and machine.

[9]
 2. SYSTEM ARCHITECTURE

Figure 1.1 System Architecture

A block diagram is a diagram of a system in which the principal parts or functions are
represented by blocks connected by lines that show the relationships of the blocks. It may
also show how the system operates, what are its inputs and outputs at various stages, and
how the information, and/or materials flow through it. The block diagram for” Online
chatting system for college enquiry knowledgeable Database” The proposed system has a
client server architecture. All the information will be kept in an optimized database on the
central server. This information can be accessed by the users through the android
application installed on their smartphones (client machines). Each client machine will have
an improved user interface.
A chatbot is a technology that allows users to have natural conversations to access content
and services. Chatbots typically take the form of a chat client, leveraging

[1
natural language processing to conduct a conversation with the user. Chatbots control
conversation flow based on the context of the users requests and respond with natural
language phrases to provide direct answers, request additional information or recommend
actions that can be taken. The diagram below provides a high-level description of how a
chat client could be used to leverage natural language processing to assist with access to
content or perform data queries.

Modules Client-Server (chat user): The proposed system has a client server architecture.
All the information will be kept in an optimized database on the central server. This
information can be accessed by the users through the android application installed on their
smartphones (client machines). Each client machine will have an improved user interface.

Chatbot: A chatbot is a technology that allows users to have natural conversations to access
content and services. Chatbots typically take the form of a chat client, leveraging natural
language processing to conduct a conversation with the user. Chatbots control conversation
flow based on the context of the users requests and respond with natural language phrases
to provide direct answers, request additional information or recommend actions that can be
taken.

[1
 3. Setting up the Environment
3.1 Overview of Required Tools and Libraries

In-depth analysis of the libraries used:

 OpenAI API: A deep dive into its functionality, capabilities, and the models it offers.
 SpeechRecognition: Examine the technical underpinnings of this library and how it interfaces
with external APIs like Google Web Speech.
 pyttsx3: Explore how text-to-speech (TTS) systems work offline, differences from online
alternatives, and its customization features.

3.2 Installation and Configuration

Step-by-step guide to setting up the environment, including potential challenges and resolutions:

 Managing API keys for OpenAI.

 Configuring microphone settings for speech recognition.
 Handling cross-platform differences in speech-to-text libraries.

3.3 Exploring OpenAI API Key Management

Explain how API keys work and the importance of safeguarding them. Discuss quotas, rate limits, and
the process of upgrading to higher usage tiers if needed

[1
 4. Speech Recognition and Voice Input
4.1 Introduction to Speech Recognition Technology

Discuss the history and advancements in speech recognition technology, starting from early systems
to modern cloud-based APIs like Google’s Web Speech.

4.2 The SpeechRecognition Library: How It Works

Explain how speech_recognition leverages multiple APIs and offers an abstraction layer for speech-
to-text conversion.

4.3 Working with Microphone Inputs

In-depth exploration of how Python interfaces with hardware (microphone). Include details on:

Setting up microphone devices.

Handling background noise.

Improving speech recognition accuracy.

4.4 Error Handling in SpeechRecognition

Advanced techniques for managing errors beyond the standard exceptions provided in the original
code. Explore retries, fallbacks to other speech recognition APIs, and more robust error reporting.

4.5 Limitations and Enhancements of SpeechRecognition

Discussion on the limitations of this library (e.g., reliance on the internet for cloud-based recognition)
and potential alternatives for offline speech recognition systems.

[1
 5. Text Generation Using OpenAI API
5.1 History and Evolution of GPT Models

Explore the development of GPT (Generative Pretrained Transformer) models, starting from GPT-1
to the current state of GPT-3, and its powerful natural language generation capabilities.

5.2 OpenAI's text-davinci-003 Model: In-Depth

Detailed analysis of the text-davinci-003 model:

 Training data.
 Model architecture.
 Strengths and weaknesses compared to other models.
 Use cases and applications.

5.3 API Usage and Request Format

Breakdown of how to interact with the OpenAI API, explaining parameters such as max_tokens,
temperature, top_p, n, and stop.

5.4 Managing Model Parameters for Better Responses

An exploration of fine-tuning responses by adjusting temperature, controlling response lengths, and

generating multiple response options.

5.5 Handling Latency and API Response Times

Discuss performance considerations when interacting with the OpenAI API, including best practices
for minimizing latency and handling API timeouts gracefully.

[1
 6. Text-to-Speech (TTS) Output with pyttsx3
6.1 Overview of Text-to-Speech Technology

The history and evolution of TTS systems, explaining the various methods (concatenative TTS,
parametric synthesis, and neural network-based TTS).

6.2 Using pyttsx3 for Offline TTS

Explore the strengths of pyttsx3 for offline usage, its underlying architecture, and how it compares to
other TTS systems.

6.3 Voice Customization in pyttsx3

Explore how users can customize the voice (e.g., selecting different voices, adjusting speed, volume,
pitch) and how this enhances the user experience.

6.4 Error Handling and Robustness in TTS

Detail the common issues encountered with TTS (e.g., voice synthesis stuttering, timing issues) and
how to handle them.

[1
 7. Detailed Code Workflow
7.1 Listening for User Input

Expand on the program's loop, detailing how the microphone listens for input continuously. Explain
the efficiency and possible optimizations of using loops for voice recognition.

7.2 Speech-to-Text Workflow

Detailed analysis of how the voice input is captured, sent to the Google Web Speech API, and
returned as text. Include network flow diagrams to visualize how the request and response flow
works.

7.3 Sending the Transcribed Text to OpenAI

Provide a comprehensive explanation of the request being formed to interact with the OpenAI API,
including debugging and logging techniques to inspect request/response cycles.

7.4 User Choice Between Text and Voice Response

Discuss the flow of user interaction, where the program allows the user to choose between hearing the
AI’s response or just seeing the text output.

7.5 Looping Back for Continuous Interaction

Explore how to structure loops in Python to allow for continuous interaction, including the pros and
cons of various looping mechanisms.

[1
DATA FLOW DIAGRAMS

Level 0 Data Flow Diagram

[1
 Level 1 Data Flow Diagram

A data flow diagram (DFD) is a graphical representation of the flow of data in a system. In
the context of the chatbot system for college enquiry using a knowledgeable database, a
DFD can be used to illustrate the flow of data between the various components of the
system. The DFD can help in understanding the data inputs, processing, and outputs of the
system. The DFD for the chatbot system can be divided into four main components: the
user interface, the natural language processing engine, the knowledgeable database, and the
response generation component. The user interface component receives the input queries
from the user in natural language. The input query is then passed on to the natural language
processing engine, which processes the query and extracts relevant information such as
intent, entities, and sentiment.

[1
 USE CASE DIAGRAM

Use Case Diagram

A use case diagram is a graphical representation of the interactions between actors (users)
and the system. In the context of the chatbot system for college enquiry using a
knowledgeable database, a use case diagram can be used to identify the various use cases
or scenarios in which the system is used.The use case diagram for the chatbot system can
include the actors (users) such as prospective students, parents, and other stakeholders who
are interested in obtaining information about the college. The various use cases can include
querying information about courses, admission requirements, campus facilities, and other
related information.

[1
 CLASS DIAGRAM

Class Diagram

A class diagram is a type of UML (Unified Modeling Language) diagram that represents
the classes and their relationships in a system. In the context of the chatbot system for
college enquiry using a knowledgeable database, a class diagram can be used to represent
the various classes in the system and their relationships.The class diagram for the chatbot
system can include classes such as User, Query, Response, Natural Language Processing
Engine, Knowledgeable Database, Retrieval-based Algorithm, Rule-based Algorithm,
Machine Learning Algorithm, Hybrid Approaches, and Feedback Mechanism. Each class
can have attributes and methods that define its behavior and properties.

[2
 SEQUENCE DIAGRAM

Sequence Diagram

A sequence diagram is a type of UML (Unified Modeling Language) diagram that

represents the interactions between objects in a system over time. In the context of the
chatbot system for college enquiry using a knowledgeable database, a sequence diagram
can be used to represent the sequence of interactions between the user and the system when
making a query. The sequence diagram for the chatbot system can include the user object,
query object, response object, and the various algorithms and components of the system
such as the Natural Language Processing Engine, Knowledgeable Database, Retrieval-
based Algorithm, Rule- based Algorithm, Machine Learning Algorithm, Hybrid
Approaches, and Feedback Mechanism.

[2
 COMPONENT DIAGRAM

Component Diagram

A component diagram is a type of UML (Unified Modeling Language) diagram that

represents the physical and logical components of a system and their relationships. In the
context of the chatbot system for college enquiry using a knowledgeable database, a
component diagram can be used to represent the various components of the system and
their relationships. The component diagram for the chatbot system can include components
such as User Interface, Natural Language Processing Engine, Knowledgeable Database,
Retrieval-based Algorithm, Rule- based Algorithm, Machine Learning Algorithm, Hybrid
Approaches, and Feedback Mechanism. Each component can have its own set of
interfaces, ports, and dependencies that define its behavior and interactions with other
components.

[2
 DEPLOYMENT DIAGRAM

Deployment Diagram

A deployment diagram is a type of UML (Unified Modeling Language) diagram that

represents the physical deployment of components and their relationships in a system. In
the context of the chatbot system for college enquiry using a knowledgeable database, a
deployment diagram can be used to represent the physical deployment of the various
components of the system.
The deployment diagram for the chatbot system can include nodes such as User Interface
Node, Natural Language Processing Engine Node, Knowledgeable Database Node, and
Feedback Mechanism Node. Each node can represent a physical machine or a logical
grouping of machines that host the corresponding components.

[2
COLLABORATION DIAGRAM

Collaboration Diagram

[2
STATE CHART DIAGRAM

State Chart Diagram

[2
 8. Advanced Error Handling Techniques
8.1 Managing Speech Recognition Errors

Expand on how to handle errors related to internet connectivity, API limits, and environmental
noise that may impact the accuracy of speech recognition.

8.2 Handling OpenAI API Errors

Discuss common errors such as API rate limiting, timeouts, or invalid requests. Explore how to
implement exponential backoff strategies, retries, and more robust error reporting to handle these
issues.

8.3 Ensuring TTS Stability

Explore scenarios where TTS might fail (e.g., audio driver issues, text exceeding length limits),
and how to safeguard against these failures.

8.4 Adding Logging and Debugging for Troubleshooting

Guide on implementing logging for every stage of the program (e.g., capturing the user’s speech,
sending API requests, and playing back audio). This helps in troubleshooting when things go
wrong.

[2
 9. Expanded User Interaction and Looping
9.1 Enhancing the User Experience

Discuss how the program’s UI/UX could be improved by providing better prompts, visual
feedback (e.g., text display while listening), and more user-friendly error messages.

9.2 Expanding the Command Set

Explore ways to expand the scope of the program by adding more voice commands, including
handling multi-turn conversations and context retention for more complex interactions.

9.3 Session Management and History

Discuss how to maintain session context so the program can "remember" previous interactions,
allowing for more natural and context-aware conversations with the AI.

[2
 10. Use Cases and Applications
10.1 Voice-Controlled Virtual Assistants

Provide an expanded explanation of how this type of technology can be used to create voice-
controlled virtual assistants for various industries.

10.2 Hands-Free Interaction in Healthcare

Explore the benefits of implementing this system in healthcare environments, especially for
hands-free interactions for medical professionals.

10.3 AI-Powered Learning and Education Tools

Discuss how voice recognition and AI text generation can be leveraged for creating intelligent
tutoring systems.

10.4 Industrial and Safety Applications

Expand on the potential of voice-controlled systems in high-risk environments such as

manufacturing, where hands-free interaction is critical for safety.

[2
 11. Performance Optimizations
11.1 Reducing API Latency

Explore advanced techniques for reducing latency when interacting with the OpenAI API, such
as pre-fetching responses, multi-threading, and handling large volumes of queries.

11.2 Improving Speech Recognition Accuracy

Discuss techniques for improving the accuracy of speech recognition, including optimizing
microphone settings, using noise-canceling techniques, and incorporating domain-specific
language models.

11.3 Optimizing TTS Performance

Investigate how to optimize the performance of the TTS engine, such as reducing synthesis time,
adjusting buffer sizes, and handling longer text inputs.

[2
 Conclusion
This project shows that, using the latest advances in Artificial Intelligence (AI) and Natural
Language Processing (NLP), a chatbot developer can produce a robust QA (Quality Assurance)
system that can be used to assist customers. OpenAI was selected for use because is currently
leading in the AI Natural Language Processing (NLP) field with generative-pretrained
transformer models. OpenAI does charge a small fee for the use of their Application
Programming Interfaces (APIs), but it was negligible for this project. We also chose the
LangChain framework because of its ability to consume multiple documents. Although a lot of
the processing is done behind the scenes, it is a powerful set of libraries that can be used to build
AI applications albeit with a bigger learning curve than the OpenAI API. The OpenAI version
needs some performance enhancements for a production system, as it otherwise can take too long
to produce an answer. One thing to look at is lowering the number of results returned from the
search_index function from 20. This could however possibly reduce the accuracy of the results.
Further testing would need to be done. Further research could also be done 6. RESULTS 7.
CONCLUSIONS Proceedings of the 15th International Multi-Conference on Complexity,
Informatics and Cybernetics (IMCIC 2024) 154 Proceedings of the 15th International Multi-
Conference on Complexity, Informatics and Cybernetics (IMCIC 2024) using vector stores like
Chroma rather than a JSON (JavaScript Object Notation) formatted file. The LangChain version
performs much better.

[3
 Reference
1. Rosruen, N., & Samanchuen, T. (2018). Chatbot Utilization for Medical Consultant
System. 2018 3rd Technology Innovation Management and Engineering Science
International Conference (TIMES-iCON), 1-5.
2. Karri, S. P. R., & Kumar, B. S. (2020, January). Deep learning techniques for
implementation of chatbots. In 2020 International Conference on Computer
Communication and Informatics (ICCCI) (pp. 1-5). IEEE.
3. Manning, C. D., Surdeanu, M., Bauer, J., Finkel, J. R., Bethard, S., & McClosky, D.
(2014, June). The Stanford CoreNLP natural language processing toolkit. In Proceedings
of 52nd annual meeting of the association for computational linguistics: system
demonstrations (pp. 55-60).
4. Plisson, J., Lavrac, N., & Mladenic, D. (2004, May). A rule based approach to word
lemmatization. In Proceedings of IS (Vol. 3, pp. 83-86).
5. Chuan, C. H., & Morgan, S. (2020). Creating and Evaluating Chatbots as Eligibility
Assistants for Clinical Trials: An Active Deep Learning Approach towards User-centered
Classification. ACM Transactions on Computing for Healthcare, 2(1), 1-19
6. Nuruzzaman, M., & Hussain, O. K. (2018, October). A survey on chatbot
implementation in customer service industry through deep neural networks. In 2018 IEEE
15th International Conference on e-Business Engineering (ICEBE) (pp. 54-61). IEEE.
7. Kapočiūtė-Dzikienė, J. (2020). Intent Detection Based Lithuanian Chatbot Created via
Automatic DNN Hyper-Parameter Optimization. Frontiers in Artificial Intelligence and
Applications, 328, 95 102.
8. Hristidis, V. (2018, September). Chatbot technologies and challenges. In 2018 First
International Conference on Artificial Intelligence for Industries (AI4I) (pp. 126-126).
IEEE.
9. Lister, K., Coughlan, T., Iniesto, F., Freear, N., & Devine, P. (2020, April). Accessible
conversational user interfaces: considerations for design. In Proceedings of the 17th
International Web for All Conference (pp. 1-11).
10. Cuayáhuitl, H., Lee, D., Ryu, S., Cho, Y., Choi, S., Indurthi, S., ... & Kim, J. (2019).
Ensemble based deep reinforcement learning for chatbots. Neurocomputing, 366, 118-
130.
11. Ng, C., & Chua, A. (2020). Training of a deep learning algorithm for quadcopter
gesture recognition. International Journal of Advanced Trends in Computer Science and
Engineering, 9(1), 211.
12. Leong, P. H., Goh, O. S., & Kumar, Y. J. (2020). Deep Learning for Conversational
Agent via Context Question Answering Model. International Journal, 9(5).

[3