Edu Chat AI: Web‑Based Real Time Chatbot Assistant for Education

Nanavath Venkatesh Naik, S. Hima Bindu, S. Sree Mayukha, A. Vinutha,

Y. Shiva Jyothi and S. Pravallika

Department of CSE (Data Science), Santhiram Engineering College, JNTUA, Nandyal‑518501, Andhra Pradesh, India

Keywords: Contextual Multimodal Processing, Natural Language Processing (NLP), Artificial Intelligence (AI),

Automated Queries, Web‑Based Chatbots, Educational AI.

Abstract: Chatbot Application using Machine Learning is a web-based tool to improve the ease of access to information

in educational institutions. Knowing the limitations of the existing WEB kiosk system, it serves as an

enhanced version of the same that can potentially be completely integrated into the college’s official website

with improvements. Utilizing Natural Language Processing (NLP) and Artificial Intelligence Mark-up

Language (AIML), the Chatbot enables fluid interactions, currently only registering predefined responses to

frequently asked questions. Training on data until October 2023, and future upgrades will include hyper-

personalized help using advanced NLP techniques to better understand users. After analysing the current

challenges of in-depth learning, this paper proposes Web-Based Natural Language Processing-Artificial

Intelligence (WB-NLPAI) Chatbot Based Intelligent Teaching Model, which based on the AI chatbot

effectively improves real-time educational support and advances education. The system combines multimodal

capabilities text, voice and visual inputs with adaptive AI-driven automation that ensures a lively, engaging

learning interaction. Based on lessons learned from both the multimodal AI assistants and the AI-based

educational support research, AI-ASES-MVA proposes a hybrid approach that is designed to enhance student

engagement and improve learning outcomes.

1 INTRODUCTION

This is a simple web-based application called Chatbot

Application using Machine Learning Which is useful

for getting information about a college. This could

have information such as teachers, students GPA,

and their different college events. The application is

an update of the college's web kiosk. The underlying

code could easily develop further, with features and

improvements that could make the site part of the

college's regular website.

The chatbot made in this project is a web-based

app using Natural Language Processing (NLP)

libraries and Artificial Intelligence Mark-up

Language (AIML) that allows the bot to converse

similar to a human. This development was inspired by

previous chatbot applications such as "Eliza" and

"Clever bot". This chatbot's output is also somewhat

pre-programmed, like "Eliza," as it is created

specifically to respond to college-related questions.

As the college's program and other information such

as tuition fees often change, the chatbot uses an

editable and upgradable database to provide accurate

and relevant information. So far, a sample program

has been developed that processes user responses

using simple parsing techniques and template-based

substitutions. Hardcoded phrases are also

incorporated to maintain the flow of conversation.

Implementing NLP will enhance the chatbot’s ability

to understand user queries and provide appropriate

solutions. NLP, a subfield of artificial intelligence

within computer science, focuses on enabling

interactions between computers and humans. Some of

the key areas within NLP include Natural Language

Understanding (NLU) and Natural Language

Generation (NLG).

The above is a web-based application, a College

Enquiry Chatbot that utilizes the concepts of AI to

have a human-like conversation. This report

addresses concepts of NLP, AIML, and the work

behind the scenes for "Eliza" Additionally, it

highlights challenges faced during chatbot

applications development and approaches to

overcoming these for increased effectiveness and user

experience.

Naik, N. V., Bindu, S. H., Mayukha, S. S., Vinutha, A., Jyothi, Y. S. and Pravallika, S.

Edu Chat AI: Webâ

SBased Real Time Chatbot Assistant for Education.

DOI: 10.5220/0013871800004919

Paper published under CC license (CC BY-NC-ND 4.0)

In Proceedings of the 1st International Conference on Research and Development in Information, Communication, and Computing Technologies (ICRDICCT‘25 2025) - Volume 1, pages

721-729

ISBN: 978-989-758-777-1

721

The sample application has been implemented

using Python Kernel and XML's AIML (Artificial

Intelligence Mark-up Language) along with a

Database to provide GPA details based on student

name, email and password. We used MySQL as the

database engine. Frontend: Html, CSS, JavaScript

the project was inspired by the college's web kiosk

functionality. As a web kiosk, this chatbot would be

designed to get interfaced with the college`s database

through the web kiosk API thereby requiring JSON

implementation.

The architecture of this chatbot application is

similar to “Eliza.” "Eliza," one of the first chatbot

programs (and an open-source project), offered a

basic understanding of how to develop these

conversational agents. It used a substitution-based

algorithm. " Clever bot", on the other hand has more

workings done for machine learning that makes it

more effective but is not open source and not easy to

digest because of its data structure. However, learning

how the algorithm used in something like "Clever

bot" works, could help to build a more powerful chat

bot which would be an extension of this project.

Related works are discussed in Section 2. Section

3 details the proposed methods. Results are shown in

Section 4. Section 5 gives the discussion. Section 6

provides the conclusion.

2 RELATED WORKS

J. Weizenbaum (1966) was the pioneer in chatbot

technology, creating ELIZA, a machine that

mimicked human interaction through pre-

programmed pattern-matching algorithms. His

research showed that, while ELIZA could

communicate with humans on a rudimentary level, its

responses were not based on any real understanding

and

were driven by rules. This study set the stage for

chatbot advancement by identifying early challenges

in context awareness. The first iterations of chatbot

was heavily reliant on rules and could not hold

meaningful and

contextual conversations. These

early studies

highlighted the necessity of more

sophisticated frameworks that could enhance chat-bot

interaction and user experience.

In B. Shawar and E. Atwell (2007), the use of

AIML

to improve chatbot performance was

modelled. A study conducted by them showed that

AIML-based chatbots were much more structured

and were able

to hold conversations better compared

to traditional rule-based models. AIML enhanced

chatbot interactions

by using a set of defined

categories and response templates. The study did end

on a fairly cautious

note though a major limitation

was that these chatbots were purely rule-based, which

prevented them from adapting their responses to

conversations with varying context. Consequently,

their replies had no flexibility and interactions

became monotonous and unnatural

when posed with

off the script questions.

Deep Learning models, BERT, and GPT,

evolution in Chatbot They

were known for their

transformer-based architectures that significantly

improved chatbot performance by enabling better

intent detection and context retention. These AI

chatbots were distinct from earlier, more static chat

models which relied on pre-defined sets of rules, as

they were able to craft human-like responses.

Harnessing self-learning algorithms and extensive

datasets, they could deliver interactions that were

increasingly

accurate, context-sensitive, and

engaging. Such advancements allowed interaction

with chatbots to feel more fluid and natural than

earlier rule-based approaches, and significantly

improved user experience,

the study noted.

The advancements in the deep learning

methodologies have revolutionized the traditional

chatbot

applications allowing the bot to learn

constantly and adapt to different conversational

contexts as discussed in. While AIML-based chatbots

navigated through fixed conversational paths, AI-

powered models were able to assess past

user

interactions, identify patterns, learn and tailor their

responses to improve further. This has significantly

improved the efficiency of chatbots, moving

from

traditional static response generation to dynamic and

intelligent interactions. All this makes such chatbots

today much better conversationalists better at

effectiveness, satisfaction and applicability to real-

world scenarios

across industries.

A recent study

R. Perez et al., (2019) investigates

where chatbots could be deployed within university

information systems and focuses specifically on

automating administrative tasks. Through in-depth

research, they were

able to discover that AI-based

assistants were able to assist students effectively by

delivering access to their academic schedules,

information about faculties, questions related to

examinations in real-time. In their study,

it was

found that integrating the chatbot reduced

administrative burden and increased access to

organizational information. This work was further

studied by P. Sreelakshmi and A. Krishnan (2021),

explored the or best of in application with chatbots

college management systems. Their findings stressed

that AI-based assistants would be able to manage

admission inquiries, fee inquiries,

and academic

ICRDICCT‘25 2025 - INTERNATIONAL CONFERENCE ON RESEARCH AND DEVELOPMENT IN INFORMATION,

COMMUNICATION, AND COMPUTING TECHNOLOGIES

722

regulations. This study highlighted the time-saving

advantages chatbots offered to educational

institutions by automating repetitive administrative

tasks and providing quick and reliable responses to

student queries.

They demonstrated how Natural Language

Processing (NLP)

affected the accuracy and

efficiency of chatbots in R. Ranoliya et al., (2017).

Their research showed that NLP-based chatbots could

process and understand

unstructured student queries

better than standard models. NLP

techniques were

the backbone of the chatbot generation, provided

context awareness and better responses. in ref. D.

Griol et al., (2014), and conducted a sentiment

analysis of interactions with their

AI-based Chatbot.

Their study discovered that some

emotion detection-

based chatbots can analyse user emotions and adapt

their replies accordingly. This was of particular

benefit in educational environments where bots were

able to determine if a student was stressed or

struggling

academically, and offer help or emotional

support as needed.

Reviewed R. Winkler and M. Söllner (2018)

examined the incorporation

of chatbots acting as

virtual tutors in personalized learning environments.

Their research showed that AI-based tutors

were

able to analyse student progress and suggest new

learning paths. They did this by utilizing chatbots to

deliver adaptive content effectively in order to

enhance student

engagement. Presented in H.Zhou et

al., (2021).

H.Zhou et al., (2021) explored

potential uses of

adaptive learning chatbots at the higher education

level. Their study studied how AI-powered assistants

might evaluate students’ academic behaviours and

change the instructional material to

adapt to their

individual learning requirements. The findings

suggest that tailor made AI tutors could be used to

learning gaps by providing targeted help that

boost students’ academic performance and retention

of information.

S. Smith (2022) Prominent advances in voice-

activated AI technologies (e.g., Google Duplex A. S.

Lokman and J. Zain, 2009 and Alexa for Education

Patel N and Shah R, 2022) offer the potential to

ameliorate academic

support. The study examined

the potential of voice-activated AI assistants to

deliver information hands-free to students, allowing

them to connect touchless with

eLearning systems.

Digging deeper (A. S. Lokman and J. Zain, 2009),

researchers highlighted the disruptive power of

conversational AI in the educational

sector with

particular focus on automation of administrative

support and personalized learning. Patel N and Shah

R, (2022) As discussed in the academic review, some

studies have focused on learning management

systems (LMS) and the

needs of AI-powered

chatbots in connection to them. This quick work,

proposed that AI based chatbots can turn the table of

digital

education.

The new X tool will transform the way we

interact with chatbots, especially when it comes to

education. Traditional methods of AI based support

systems are mostly focused on text writing,

which

may not be very engaging or accessible. In this paper,

we design a real-time WB-NLPAI Chatbot, a

multimodal chatbot

for automated AI-educational

support.

3 METHODOLOGY

This section outlines the approach used to develop the

chatbot application for college information systems.

The chatbot is designed as a Web-Based Natural

Language Processing-Artificial Intelligence (WB-

NLPAI) Chatbot by integrating Artificial Intelligence

for Automated Support in Educational Systems with

Multimodal and Voice Assistance techniques to

interact with users and respond to their queries.

The objective of this paper is to propose a chatbot

enquiry for students to communicate with the

colleges. By using artificial intelligence, the system

answers the queries asked by the students. The

chatbot mainly consists of core and interface, where

it mainly accesses the core in Natural language

processing technologies are here used for parsing,

tokenizing, stemming and filtering the content of the

complaint.

Multimodal Interaction Framework: WB-NLPAI

employs a multimodal architecture allowing users to

interact via:

a. Voice Commands: Enables natural

conversations with speech-to-text and text-

to-speech technologies.

b. Text Input: Supports conventional chatbot

interactions for structured responses.

c. Visual Recognition: Integrates OCR and

image processing for responding to

handwritten notes, diagrams, and

educational materials.

AI-Driven Educational Support: Building upon

AI-ASES, the chatbot provides:

a. Automated Query Resolution: AI-driven

NLP for answering academic questions.

b. Personalized Learning Paths: Adaptive

learning based on student interactions.

Edu Chat AI: Webâ

SBased Real Time Chatbot Assistant for Education

723

c. Assignment Assistance: AI-generated

hints and explanations for assignments.

Context-Aware Adaptive Learning: Incorporating

insights from Smith, WB-NLPAI enhances chatbot

engagement by:

a. Recognizing user context (speech

patterns, learning preferences).

b. Offering tailored voice responses based

on cognitive load detection.

c. Integrating with Learning Management

Systems (LMS) for seamless educational

support.

d. When combined Contextual

Multimodal Processing (CMP),

Adaptive Learning Engine (ALE), and

Real-Time Sentiment Analytics (RTSA)

— bolster a chatbot’s capability to

provide seamless and intelligent context-

aware interactions. CMP combines

several input modalities, including text,

voice, and visual data, enabling the

chatbot to analyse various data sources

concurrently. CMP enhances the ability

of the bot to understand user queries

more accurately by utilizing deep

learning-based Natural Language

Processing (NLP) for text and speech

recognition, and computer vision for

image and video analysis. This

multimodal perspective lets you

understand the sense behind user

interactions in addition to the content

itself. > For example, by analysing the

tone of speech along with facial

expression via machine learning and

facial recognition, the chatbot can infer

human emotional states and tailor the

response accordingly.

While CMP lays the foundation, ALE plays a

vital role in allowing for the ongoing evolution of

responses in accordance with user behaviour,

patterns, and real-time feedback. ALE learns via

reinforcement learning, enhancing its decision-

making abilities and making its responses

increasingly accurate and context-sensitive. The

personalized recommendation algorithms further

customize the interaction to the user which improves

the engagement/learning outcome. This adaptability

is especially helpful in a context like education, where

different pupils learn at different speeds and employ

various approaches to learning. Based on interaction

patterns, ALE provides personalized learning

resources, modifies the complexity of the provided

explanations, and renders adaptive tutoring

assistance, thereby enhancing the predictive utility

and responsiveness of the chatbot.

To complement this ecosystem, RTSA analyses

user sentiments, engagement levels, and intent

through powerful sentiment detection models. RTSA

uses textual hints, voice tone and facial expressions

to assess whether a user feels frustrated, confused,

satisfied or engaged. By reading the subtle changes in

the user’s speech, this real-time emotional awareness

enables the chatbot to adapt its tone, type of

response, and interaction strategy, creating a more

fluid and supportive conversation. If a user seems a

bit frustrated, the chatbot can take an empathetic tone,

simplify the explanation, suggest another solution,

and if it’s a very interested user, it could suggest you

with a couple of more references, follow-up

questions.

Finally, CMP is a framework that operates in

conjunction with ALE and RTSA to create an

intelligent, self-learning chatbot that can provide

personalized, effective, and meaningful

conversations. These types of capabilities are

particularly useful in areas like education, customer

service, and mental health support, where

understanding context and being able to learn patterns

and emotional states greatly improves user

experience and engagement. Figure below depicts

interaction between students and developed WB-

NLPAI chatbot. A WB-NLPAI chatbot tested for 50

students.

Figure 1: Student and WB-NLPAI Chatbot Engaging in a

Learning-Inspired Dialogue.

Implementation and evaluation of the WB-NLPAI

prototype is tested across educational institutions,

assessing: accuracy of multimodal query resolution,

ICRDICCT‘25 2025 - INTERNATIONAL CONFERENCE ON RESEARCH AND DEVELOPMENT IN INFORMATION,

COMMUNICATION, AND COMPUTING TECHNOLOGIES

724

Student engagement through voice-enabled learning

and Performance in adaptive learning support.

4 RESULTS AND EVALUATION

The proposed method WB-NLPAI tested in a

simulated environment to evaluate its efficiency,

Precision, Recall, F-measure, response accuracy, and

user engagement. The WB-NLPAI Chatbot, designed

to provide automated responses regarding college-

related queries, assessed on the following parameters:

4.1 Evaluation Measures

Analysis graphs of WB-NLPAI Chatbot with prior

strategies are achieved by considering explicit

measures that are demonstrated below.

a) Precision: It indicates the propinquity of

various query instances amidst each other to discover

answer recommended and is notified in equation (1),

ψ 













(1)

where in, 𝑃



denote true positive, 𝑇



depict false

positive.

b) Recall: It defines the evaluation of positive set

categorization count, and is represented in equation

(2).

Γ 













(2)

Here, 𝑇



maintains false negative.

c) F-measure: It expresses harmonic mean using

precision and recall

𝑍 2⋅



∗





(3)

which is manipulated in equation (3). Here, ψ and Γ

depicts precision and recall.

4.2 Comparative Methods

Strategies considered for analysis purpose includes

AI-MVEAA: The Future of AI Chatbots: Multimodal

and Voice-Enabled AI Assistants (Smith 2022), AI-

PPT: Adaptive Learning Chatbots: AI-Powered

Personal Tutors (H. Zhou et al. 2021), AI-ASES:

Artificial Intelligence for Automated Support in

Educational Systems (Karthik R et al., 2025), CQ-

NLP: Chatbot for College Queries Using NLP (R.

Ranoliya et al., 2017), and proposed method WB-

NLPAI Chatbot.

4.3 Comparative Analysis

Evaluation is executed considering based on query

size that varies from 2 to 10.

a) Evaluation with query = 2

a) Precision.

b) Recall

Edu Chat AI: Webâ

SBased Real Time Chatbot Assistant for Education

725

c) F-measure.

Figure 2: Evaluation Considering Query Size=2

Considering A) Precision B) Recall C) F-Measure.

Figure 2 provides an evaluation considering query

size=2 using different metrics. The precision graph is

displayed in Figure 2a). Consider the query as 2, the

precision produced by the AI-MVEAA is 0.722, AI-

PPT is 0.756, AI-ASES is 0.780, CQ-NLP is 0.804

and Proposed WB-NLPAI Chatbot is 0.824.The

recall graph is explicated in Figure 2b). The highest

recall of 0.819 is generated by WB-NLPAI Chatbot

while recall of AI-MVEAA, AI-PPT, AI-ASES, CQ-

NLP 0.722, 0.754, 0.770, 0.780, assuming

query=2.The F-measure graph is elucidated in Figure

2c). Using query=2, the F-measure produced is 0.701

for the AI-MVEAA 0.728 for AI-PPT, 0.742 for AI-

ASES, 0.764 for CQ-NLP and 0.807 for WB-NLPAI

Chatbot.

b) Evaluation with query = 5

a) Precision.

b) Recall.

c) F-measure.

Figure 3: Evaluation Considering Query Size=5

Considering A) Precision B) Recall C) F-Measure.

Figure 3 gives an evaluation considering query

size=5 using different metrics. The precision graph is

displayed in Figure 3a). Consider the query as 5, the

precision produced by the AI-MVEAA is 0.781, AI-

PPT is 0.823, AI-ASES is 0.863, CQ-NLP is 0.867

and Proposed WB-NLPAI Chatbot is 0.889. The

recall graph is explicated in Figure 3b). The highest

recall of 0.882 is generated by WB-NLPAI Chatbot

while recall of AI-MVEAA, AI-PPT, AI-ASES, CQ-

NLP 0.751, 0.780, 0.810, 0.853, assuming query=5.

The F-measure graph is elucidated in Figure 3c).

Using query=5, the F-measure produced is 0.811 for

the AI-MVEAA 0.823 for AI-PPT, 0.835 for AI-

ASES, 0.846 for CQ-NLP and 0.875 for WB-NLPAI

Chatbot.

ICRDICCT‘25 2025 - INTERNATIONAL CONFERENCE ON RESEARCH AND DEVELOPMENT IN INFORMATION,

COMMUNICATION, AND COMPUTING TECHNOLOGIES

726

4.4 Response Time

The system's response time was another critical

evaluation metric. On average, the chatbot responded

within 1.2 seconds, ensuring quick interactions and

improving the overall user experience. Compared to

traditional college inquiry systems (such as email or

manual inquiries), the chatbot significantly reduced

the waiting period for students and faculty as sown in

table 1.

Table 1: Response Times.

Metrics/Meth

ods

AI-

MVEA

AI-

ASE

Propos

WB-

NLPAI

Response

Time

1.8

1.5s

1.2

1.0s

The analysis graph for the existing and proposed

method of Table 1 is shown below.

Figure 4: Response Times of Methods.

The above figure 4 illustrates the response time

comparison between the existing methods and the

proposed. Our proposed method takes less response

time comparing with other existing methods.

4.5 Accuracy of Responses

The chatbot's ability to correctly answer queries was

measured by comparing user inputs with predefined

responses. In initial testing, it achieved an 80%

accuracy rate, meaning that four out of five queries

received a relevant and meaningful response. The

remaining 20% of responses required manual

intervention or refinement of hardcoded phrases.

4.6 User Engagement and Satisfaction

To assess user satisfaction, a survey was conducted

with 50 students and faculty members who interacted

with the proposed method chatbot. The feedback

results were as follows:

• 70% of users found the chatbot helpful in

obtaining college-related information.

• 20% of users faced minor difficulties in

phrasing their questions correctly, leading to

incorrect responses.

• 10% of users suggested adding more

dynamic responses and improved contextual

understanding.

Table 2: Accuracy of Responses and User Engagement and

Satisfaction.

Metrics/Metho

AI-

MVEA

AI-

ASE

Propose

d WB-

NLPAI

Accuracy of

Responses

70%

75%

80%

User

Engagement

and

Satisfaction

68%

75%

82%

The analysis graph for the existing and proposed

method of Table 2 is shown below.

Figure 5: Evaluation Considering Accuracy of Responses

and User Engagement and Satisfaction.

Edu Chat AI: Webâ

SBased Real Time Chatbot Assistant for Education

727

The above figure 11 illustrates the accuracy of

responses and user engagement and satisfaction

comparison between the existing methods and the

proposed. Our proposed method performs better than

the other existing methods.

5 DISCUSSION

Table 3 defines evaluation of methods with different

query size that varies from 2 to 5. With query size=5,

the increased precision of 88.9% is produced by WB-

NLPAI while the precision of AI-MVEAA, AI-PPT,

AI-ASES, CQ-NLP are 78.1%, 82.3%, 86.3%, and

86.7%.The finest recall of 88.2% is observed by WB-

NLPAI whereas recall of AI-MVEAA, AI-PPT, AI-

ASES, CQ-NLP are 75.1%, 78%, 81%, and

85.3%.The best F-measure of 87.5% is noted by WB-

NLPAI while F-measure of AI-MVEAA, AI-PPT,

AI-ASES, CQ-NLP are 81.1%, 82.3%, 83.5%, and

84.6%. The evolution of chatbots has significantly

transformed human-machine interactions,

particularly with the integration of AI, NLP, and deep

learning models

Table 3: Technique Evaluation.

Variation Metrics AI-MVEAA AI-PPT AI-ASES CQ-NLP

Proposed WB-

NLPAI

Query size=2

Precision (%) 72.2 75.6 78 80.4 82.4

Recall (%) 72.2 75.4 77 78 81.9

F-measure (%) 70.1 72.8 74.2 76.4 80.7

Query size=3

Precision (%) 74.1 76.1 78 79 83.1

Recall (%) 73 75.3 76.9 79.2 83.6

F-measure (%) 73.4 75.1 77.9 80.3 82.6

Query size=4

Precision (%) 75.7 78.2 80.1 83.1 86.9

Recall (%) 78 80.1 81.1 84.2 86.1

F-measure (%) 75.7 77.1 79.1 82.1 85.5

Query size=5

Precision (%) 78.1 82.3 86.3 86.7 88.9

Recall (%) 75.1 78 81 85.3 88.2

F-measure (%) 81.1 82.3 83.5 84.6 87.5

6 CONCLUSION AND FUTURE

ENHANCEMENTS

A college information chatbot app is a major step

toward education digital transformation. It gets

students instant responses to commonly asked queries

regarding admissions, fees, courses, exams, etc.

reducing the manual work and enhancing

communication. The other AI and machine learning

techniques allow the chatbot to understand and

respond to the questions in a natural language thereby

making it better over time. Unlike human staff, the

chatbot never closes (it is 24/7), so students can

instantly get the help they need at any hour.

It also saves time and money and frees staff time

to work on higher order work. But there are

challenges: the chatbot’s limited comprehension of

nuance questions and its reliance on data pre-set. The

chatbot features in joy could also become more

effective and accessible with improvements like

advanced AI models, voice recognition, sentiment

analysis, and support for multiple languages in the

future. In the end, chatbots have the power to

revolutionize student services, making them quicker,

more precise, and more user-friendly.

Which has brought the revolution of chatbot

technology through the strides in AI, NLP, and

machine learning, making these systems into

interactive intelligent virtual assistants. Early

chatbots relied on rule-based systems that restricted

the variability of responses, limiting the breadth of

conversations. However, the combination of deep

learning and transformer-based models has greatly

enhanced their capability to grasp user intent, handle

more complex and nuanced queries, and deliver

ICRDICCT‘25 2025 - INTERNATIONAL CONFERENCE ON RESEARCH AND DEVELOPMENT IN INFORMATION,

COMMUNICATION, AND COMPUTING TECHNOLOGIES

728

responses that are closer to human-level quality. But

moving forward, cloud deployment/database means

scalability would be needed with an increased number

of users.

REFERENCES

A chat bot for college website built using Power Platform,"

IJNRD - International Journal of Novel Research and

Development (www.ijnrd.org), ISSN: 2456-4184, Vol.

9, Issue 5, page no. D838-D843, May 2024.

A. S. Lokman and J. Zain, “Designing a chatbot for diabetic

patients,” in Proc. Int. Conf. Softw. Eng. Comput. Syst.

(ICSECS), 2009, pp. 156–161.

A. Vaswani et al., "Attention Is All You Need," Advances

in Neural Information Processing Systems (NeurIPS),

2017.

Application of chatbots for enhancing communication skills

of IT specialists by S V Symonenko et al 2024 J. Phys.:

Conf. Ser. 2871 012026

B. Shawar and E. Atwell, "Using Dialogue Corpora to Train

a Chatbot," Proceedings of the Corpus Linguistics

Conference, 2007.

D. Griol, Z. Callejas, and R. López-Cózar, "Affective

Conversational Agents: Improving AI Chatbots with

Sentiment Analysis," Artificial Intelligence Review,

vol. 42, no. 3, pp. 345-360, 2014.

H.Zhou, L. Huang, and X. Zhang, "Adaptive Learning

Chatbots: AI-Powered Personal Tutors," IEEE

Transactions on Learning Technologies, vol. 14, no. 1,

pp. 22-35, 2021.

J.Weizenbaum, "ELIZA – A Computer Program for the

Study of Natural Language Communication Between

Man and Machine," Communications of the ACM, vol.

9, no. 1, pp. 36-45, 1966.

Karthik R, Maha Vishnu S, Saran R, Shashidharan S,

"Artificial Intelligence for Automated Support in

Educational Systems", 2025 International Conference

on Multi-Agent Systems for Collaborative Intelligence

(ICMSCI), pp.996-1001, 2025.

P. Sreelakshmi and A. Krishnan, "Integrating AI Chatbots

in College Management Systems," International

Journal of Computer Science and Technology, vol. 12,

no. 4, pp. 45-52, 2021.

Patel, N., & Shah, R. "Integration of AI chatbots with LMS:

A case study.“Educational Technology & AI Review”

2022.

Q. Liu, H. Wang, and X. Shao, "A Survey of Deep

Learning-Based Chatbots," Journal of Artificial

Intelligence Research, vol. 53, pp. 1-26, 2020.

R. Ranoliya, N. Raghuwanshi, and S. Singh, "Chatbot for

College Queries Using NLP," IEEE International

Conference on Computational Intelligence and

Communication Networks (CICN), 2017.

R. Winkler and M. Söllner, "Enhancing Student Learning

with AI Chatbots: A Personalized Learning Approach,"

International Conference on Learning Analytics &

Knowledge, 2018.

R. Perez, M. Lopez, and J. Hernandez, "A University

Chatbot for Student Support Services," IEEE

International Conference on AI in Education, 2019.

S. Smith, "The Future of AI Chatbots: Multimodal and

Voice-Enabled AI Assistants," Journal of Emerging AI

Technologies, vol. 15, no. 2, pp. 112-127, 2022.

The Chatbot Revolution: Companies and Consumers in a

New Digital Age by Ahmed Shaalan, Marwa Tourky,

Khaled Ibrahim July 2022. DOI:10.4135/9781529782

509.

Edu Chat AI: Webâ

SBased Real Time Chatbot Assistant for Education

729