Chatbot for Student Enquiry by IRJET Journal

Chatbot for Student Enquiry

Pooja Kandari1 , Rakshitha S2, Samatha K3 , Samreen Sultan L B4, Shwetha Shree A5

1,2,3,4

UG Student, Dept. of Computer Science and Engineering, Ballari Institute of Technology and Management, Ballari, Karnataka, India

5 Assistant Professor, Dept. of Computer Science and Engineering, Ballari Institute of Technology and Management, Ballari, Karnataka, India

Abstract – This chatbot has been designed to assist users in resolving queries related to the college, cateringspecifically to the needs of new students and their parents, who often have numerous questions about the admission process, fee structure, and facilities offered by the institution. By automating the enquiry process, the chatbot significantly reduces the manual effort and time spent by both the college staff and prospective students or their parents in addressing these concerns. To enhance user experience and accessibility, the bot is equipped with advanced features such as speech recognition, text-to-speech, and multilingual support. The speech recognition capability allows users to provide voice commands instead of typing, making interactions more convenient and user-friendly. Additionally, the text-to-speech functionality enables users to hear responses to their queries, further simplifying the process for individuals who may find reading difficult or inconvenient. The multilingual support feature ensures that the bot can communicate effectively in two widely used local languages, Telugu and Kannada, in addition to English, thereby making it accessible to a broader audience. The primary goal of this chatbot is tostreamline and automate the operations at the enquiry desk, reducinghuman workload while ensuring prompt and accurate responses to user queries. This system not only improves efficiency but also enhances the overall experience for users seeking information, ensuring a hassle-free interaction process for allstakeholders.

Key Words: Natural Language Processing, Speech Recognition,Chatbot,Text-to-Speech,PatternMatching,TFIDF,CosineSimilarity,Spellchecker

1.INTRODUCTION

Achatbotisacomputersoftwarethat,whenconversedwith throughtextorvoice,repliesasifitwereacleverentitythat understands one or more human languages using Natural LanguageProcessing(NLP)[1].Achatbot’sprimarytaskis to help users by providing answers to their questions by understandingwhathumanwantsandguidesthemtotheir desiredoutcome[2].Chatbotshaveemergedasimportant tools in the digital age, they enable smooth interaction betweenusersandsystems.

In the college context, AI-based voice responders have become integral parts of student services, assisting with tasks like course registration, academic advising, and campus navigation. They offer round-the-clock support,

freeingupadministrativestafftofocusonmorespecialized tasks while providing students with instant access to informationandassistance[3] Peoplepreferthesechatbots tolookingthingsupontheinternetbythemselvesasthese can easily provide a thorough, detailed answer to the questions that the users want answered. These chatbots provetohavetheabilitytosimulatethecognitiveabilities that humans have [4] Moreover, with the increasing popularity of smart speakers and virtual assistants like Amazon Alexa and Google Assistant, colleges have begun integrating AI-based voice responders into their communication channels, offering students another convenientwaytoaccesscollegeservicesandresources[3].

Conversational interfaces are platforms that can have conversation like a real human. Generally, users use Graphical user interfaces (GUI) to give commands to the computer.Thecomputertheninterpretsthemeaningofthat command and perform the desired action. In case of Conversational interfaces, theusercancommunicate with computer in their natural language instead of giving command or using GUI [5]. A chatbot is a software applicationusedtoconductanonlinechatconversationvia textortext-to-speech,inlieuofprovidingdirectcontactwith alivehumanagent[6]

College Chatbots are becoming increasingly popular as a means of providing quick access to helpful information. College Chatbots are used to answer common questions, provideguidance,andofferresources.Additionally,College Chatbotscanbeusedtoimprovestudentengagementand increasestudentsatisfaction[7].Chatbotsarethesourceof answers to the users’ questions in any particular domain whereitisoperating[8].Thegoalofthischatbotistoassist userswithcollege-relatedinquiries.ThechatbotusesNLP techniquesandarelationaldatabasetoprovideuserswith accurate,context-sensitiveresponsesinreal-time.Thisbot helps in handling versatile range of queries such as admissions, fees, hostel facilities, scholarships, and other informationrelatedtotheinstitution.

In an artificial intelligent field, there are some hybrid methodsandadaptivemethodsavailablewhicharemaking systems more complex. Not only that but also there is a hybrid combination of natural language processing and intelligentsystems.Thesesystemscanlearnthemselvesand

International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056

Volume: 12 Issue: 01 | Jan 2025 www.irjet.net p-ISSN: 2395-0072

renew their knowledge by reading all electronics articles availableontheinternet[10].

Users(humans)canaskquestionstosystemliketheywould to any other human. These systems are often known as internet answering engines. In addition to the internet answering engines, currently, many applications are introducedsuchaschatter-robotorknownaschatbotwhich is often aimed at giving an automatic reply or just for entertainment[10] Thetaskoftheseapplicationsisjustto answerthequestionsbasedontheknowledgestoredinside them. Few of the methods used in this application are pattern-matching,naturallanguageprocessing,datamining. The chatbot would match the input sentence from the speakeroruserwiththatpatternexistedintheknowledge base.Eachpatternisthencomparedwiththeknowledgeof chatbot[10].

2. LITERATURE SURVEY

[1] Walaa Hassan et al. (2023) proposed an interactive chatbotforcollegeinquiries,focusingonimprovinghumancomputer interaction using Artificial Intelligence (AI) and NaturalLanguageProcessing(NLP).Thesystemisdesigned to assist university students by answering their queries accuratelyandefficientlythroughauser-friendlyinterface. Thechatbotanalyzesuserinput,interpretsthecontext,and generatesresponsesresemblinghumancommunication.The proposedarchitectureincludesintelligentagentstomanage communication, ensuring timely and precise answers to studentqueries.Thesystemwasexperimentallyvalidated, demonstratingits enforceability and efficiencyin reducing search time and enhancing the user experience on educationalplatforms.

[2]A.KousarNikhathetal.(2022)proposedanintelligent collegeinquirychatbotutilizingNaturalLanguageProcessing (NLP)andDeepLearningtechniques,specificallyLongShortTermMemory(LSTM)networks,toenhanceuserinteraction. Thesystemaimstoaddressacademicandadmission-related queries for freshers, students, faculty, and parents. By leveragingRecurrentNeuralNetworks(RNN)andNLP,the chatbot understands and analyzes user queries to provide accurate,human-likeresponses.Thisweb-basedapplication significantly reduces the need for manual intervention, offering quick resolutions to inquiries about admission processes, fee structures, departmental details, and other academicconcerns.Itisdesignedtoimprovetheefficiencyof college-relatedinteractions.

[3]PavithraNetal.(2024)highlightedthedevelopmentand functionalityofthe"CollegeCasioBot,"anAI-drivenvirtual assistant designed to streamline access to college-related information. Published in the International Journal of InnovativeResearchinScience,Engineering,andTechnology, theirstudyshowcasedthebot'sabilitytoprovidereal-time responses to queries regarding campus navigation, department details, fee structures, and event updates. By

integratingnaturallanguageprocessing(NLP)andoffering dual modes text and audio it ensures accessibility for diverse users. The research emphasized the bot’s role in reducing administrative workload and improving engagement by delivering instant, accurate information, particularly for remote users and parents unable to visit campusphysically.

[4] A. Balamurugan et al. (2024) proposed an AI-based chatbot integrated with voice assistance to cater to the increasing demand for advanced virtual assistants. The systemcombinesOpenAI’sGPT-3.5largelanguagemodelfor intelligentconversationalcapabilitieswithatext-to-speech APIforseamlessaudiointeraction.Itacceptsuserinputas audio,convertsittotext,generatesaresponseusingGPT-3.5, andthenprovidestheoutputinaudioformat.Theproposed system emphasizes user data privacy by ensuring that neither the text-to-speech API nor theGPT-3.5 API retains userdata.Thisinnovativeapproachbridgesthegapbetween traditionalchatbotsandmodernvoiceassistants,enhancing useraccessibilityandexperience.

[5] Sangeeta Kumari et al. (2020) proposed an interactive chatbotsystemtoaddressthechallengesfacedbystudents and parents during the college admission process. The systemintegratesNaturalLanguageProcessing(NLP)with text and audio-based interactions to facilitate seamless communication.Byprovidingaplatformwhereuserscanask queries in simple English, the chatbot reduces the dependency on physical enquiry desks and minimizes repetitivetasksforadmissiondepartments.Unliketraditional methods, the system offers 24/7 assistance, improving efficiencyandusersatisfaction.Itnotonlyanswersqueries but also incorporates self-learning capabilities to enhance servicequalityovertime.Thisimplementationaddressesthe need for scalability and cost-effectiveness, making it a valuable tool in academic environments. The literature highlights the system's potential to transform inquiry managementincolleges.

[6]HarshalaGawadeetal.(2020)proposedaCollegeEnquiry Chat-Botsystemdesignedtoprovideavirtualassistantfor resolving college-related queries. The system utilizes Artificial Intelligence and a database-driven approach to simulatehuman-likeinteractionsviatextortext-to-speech communication. Built with algorithms that analyze user queries, it addresses questions on admissions, fees, scholarships, timetables, and more. The chatbot leverages platforms like AIML and chat fuel to manage queries efficiently,offeringresponsesintext,images,orcardformats. The system is aimed at reducing manual effort, providing quicker resolutions, and enhancing the user experience duringtheacademicinquiryprocess.

[7]YaseenE.andSwamydossD.(2023)proposedaCollege Chatbot system developed using Python Flask to provide studentswithauser-friendlyplatformforaccessingcollegerelated information. The chatbot is designed to simulate

International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056

Volume: 12 Issue: 01 | Jan 2025 www.irjet.net p-ISSN: 2395-0072

human-like conversations, addressing queries related to admissions, academics, courses, student life, and more. Integrated with the college's system and external APIs, it deliversup-to-dateandpersonalizedresponsesusingNatural LanguageProcessing(NLP).Thesystemaimstoreducethe timestudentsspendsearchingforinformationandenhance theirexperiencebyofferingquick,accessible,andautomated assistance. This chatbot also provides administrators with full control overchatbotdata and user interactionhistory, ensuringefficiencyandscalability.

[8] Shashank Bhalotia et al. (2018) proposed a chatbot system for college inquiries, integrating AI and Natural LanguageProcessing(NLP)toenhancetheuserexperience oncollegewebsites.Thesystemefficientlyhandlesqueries relatedtoadmissions,examinationdetails,attendance,grade points, and placement activities. By utilizing a predefined knowledgebaseandsemanticsentencesimilarityalgorithms, the chatbot analyzes user input and provides accurate responses. This system addresses challenges like slow websitenavigationanddifficultyinlocatinginformationfor non-students, offering a user-friendly and time-saving solution through a conversational interface. It ensures personalizedandeffectiveassistancetousers.

[9]AmolHalvankaretal.(2024)proposedacollegeinquiry chatbot system utilizing Artificial Intelligence (AI) and Natural Language Processing (NLP) to address student queries related to admissions, academics, fees, and other activities.Thechatbotintegratesalgorithmstoanalyzeuser questions and provide appropriate responses through a conversational user interface. Designed as a web-based application,itenhancesuserexperiencebyofferingaquick, accessible,andefficientplatformtoresolveinquirieswithout requiringphysicalvisitstotheinstitution.Thesystemaimsto reducetheworkloadoncollegestaff,improveservicequality, andprovide24/7assistanceforstudentsandusersacross variousdevices.

[10]GuruswamiHiremathetal.(2020)proposedachatbot system designed specifically for the education sector to provideautomatedresponsestouserqueries.Unlikeexisting chatbots that rely solely on local databases, their system integratesbothlocalandwebdatabases,makingitscalable andhighlyinteractive.Thechatbotemploystechniquessuch as Natural Language Processing (NLP), pattern matching, machinelearning,anddataprocessingalgorithmstoenhance response accuracy and performance. The system aims to improveuserexperiencebyofferingdynamicandcontextaware answers, ensuring seamless communication and adaptabilitytonewinformation.

3. SYSTEM ARCHITECTURE

The system architecture consists of a client-server model where the frontend interacts with the backend through HTTP requests. The Flask server handles user authentication, processes user inputs through NLP

techniques, and fetches appropriate responses based on predefined intents stored in a JSON file. The system also integrateswithMySQLforuserdatamanagement.

Users interact with the chatbot through a web-based interface. The interface supports text input, speech recognition,andmultilingualfunctionality.Flaskappactsas the middleware, managing client-server communication. Processes user inputs, performs intent detection, and generates responses. Handles requests for translation, speechrecognition,anddatabasequeries.TheNLPEngine preprocessesuserinputstoremovenoiseandstandardize text. Performs intent classification using TF-IDF vectorizationandcosinesimilarity.Also,aMySQLdatabase isusedtostoreusercredentialstosupportauthentication. Andthebothaspredefineddictionariestohandlecommon regional terms Google Translator API supports dynamic multilingualinteractions.ThereareSpeech-to-textandtextto-speechfunctionalitiestoenhanceaccessibility.

Fig -1:SystemArchitecture

3.1 User Login:

Theuserlogsintothesystem.Thenthesystemverifiesthe user'scredentials.

3.2 User Input:

Theuserselectsanyoftheoption/ssuchasSelectLanguage, EnableText-to-Speech,andEnableSpeechOption.Whenthe user clicks on the select language dropdown, they will be

International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056

Volume: 12 Issue: 01 | Jan 2025 www.irjet.net p-ISSN: 2395-0072

provided three options. They can select either English, Kannada,orTelugu.

Whentheuserenablesthetext-to-speechfeature,theuser willbereceivingoutputintheformoftextaswellasspeech. Iftheuserwantstostopthetext-to-speechfeature,theycan dosobyuncheckingtheenabletext-to-speechbutton.

Oncetheuserenablesthespeech-to-textfeaturebyclicking the speak button, the user will be prompted to ask their queryvocally.Thesystemreceivestheuser'sinput.

3.3 Processing:

The system preprocesses the user's text input. And then systemconvertstheuser'stextintoavectorrepresentation usingTF-IDF.Iftheuserwouldhaveenabledspeakfeature to give input, then the system converts the user's speech inputintotext. Thesystem comparestheuser'sinputtoa database of responses using cosine similarity. The system searchesforrelevantresponsesinadatabase Thesystem checksthespellingoftheuser'sinput.

3.4 Response Generation:

The system generates a response based on the processed input.Theresponsecanbeintextorspeechformat.

3.5 Output:

Thesystemdisplaystheresponsetotheuser.

3.6 Administration:

An administrator can add and update information to the system'sdatabase.

4.METHODOLOGY

4.1 Flask Application

TheFlaskappservesasthemainbackend,managingrouting, database interactions, and chatbot logic. It handles user registration and login, connects to a MySQL database to authenticateusers,andmanagessessions.Italsoprocesses user inputs, interacts with the chatbot model to generate responses,andreturnsthemtothefrontend.

4.2 Natural Language Processing

(NLP)

NLPformsthecoreofthechatbot’sfunctionality,enabling thesystemtointerpretandprocessuserqueries.TheNLP pipeline begins with text preprocessing, which includes tokenization,lemmatization,andtheremovalofstop-words tostandardizeinputs.

Forthealgorithmtounderstandthesesentences,requireto getthewordsinasentenceandexplainthemindividuallyto our algorithm. So, you break down your sentence into its

constituent words and save them. This process is called tokenizing,andeachwordiscalledatoken[9].

Lemmatizationisatextpre-processingmethodthathelpsin naturallanguageprocessing(NLP)modelstobreakaword downtoitsrootmeaningtofindsimilarities[9]

These steps help reduce noise and ensure more accurate intent detection. A spell correction module corrects typographical errors, while predefined word dictionaries support regional language translations for Kannada and Telugu.TF-IDFvectorizationtransformsprocessedtextinto numerical representations, making it suitable for further computation,suchascosinesimilarity-basedmatching.This NLPlayerensuresthechatbotcanhandlediversephrasing andlinguisticvariationseffectively.

4.3 Chatbot Logic

Thechatbot'slogicisbuiltonarule-basedintentmatching systemenhancedwithflexibleresponsemechanisms.User inputsarematchedagainst a predefinedcorpusofintents storedinaJSONfile.Ifamatchisfoundbasedonsimilarity thresholds, the corresponding response is selected and returned.Forqueriesthatdonotfitpredefinedintents,the chatbotalsousesspellchecking,eveninthecaseofspelling errorsitprovidestheappropriateoutput.Thechatbotlogic also supports multilingual translation by using Google TranslatorAPIfordynamicresponsegenerationintheuser’s preferred language. Additional features, such as session management, ensure continuity during prolonged interactions. When the user is inactive for more than 30 secondsthechatbotpromptstheusertoentertheirquery.

4.4 MySQL Integration

MySQLisusedforhandlinguserdata,includingregistration and login credentials. The module interacts with the database to store and retrieve user information, ensuring secure authentication. It also manages error handling for database operations, such as handling duplicate entries duringregistration.

4.5 Frontend (HTML/CSS/JavaScript)

TheuserinterfaceisimplementedinHTML,styledwithCSS, andmadeinteractiveusingJavaScript.Thefrontendincludes a chat container for displaying user and bot messages, a languageselectiondropdownformultilingualsupport,and buttons for speech-to-text input as well as text-to-speech output. JavaScript handles user input events, enabling dynamic interactions with the backend Flask application. The layout ensures that the user experience remains seamless,withfeatureslikescrollablemessagehistoriesand real-time updates enhancing usability. The chat interface supports sending messages, displays bot responses dynamically,andincludesa timeoutfeaturefordisplaying systemmessagesafterinactivity

International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056

Volume: 12 Issue: 01 | Jan 2025 www.irjet.net p-ISSN: 2395-0072

4.6 Intent Definition

AJSONfiledefinesthestructureofthechatbot'sknowledge base. It includes various intents, each containing patterns (possible user inputs) and corresponding responses. The chatbot uses this file to determine the most relevant response based on the user's query. This modular setup allows for easy updates and expansions of the chatbot’s capabilities.

5. CONCLUSIONS

The chatbot projectisanintegration of cutting-edge technologiestodeliveraseamless,user-centricsolutionfor college-related inquiries. ByintegratingNLP techniques, multilingualsupport,androbustdatabasemanagement,the system effectively automates repetitive tasks, thereby enhancingefficiencyandreducingtheworkloadofhuman operators.

Thisisoneofthegreateststrengthsofthesystem:itsability tobeadaptivewithlinguisticdiversity,fromusersspeaking English to those preferring regional languages such as Kannada and Telugu Speech recognition and TTS functionality enhance the scope of its use for various demographics.

Althoughthe chatbotisreallyaccuratein intent detection and response generation,thereisstill room to improve in ambiguityhandlingoftheuser input and regional dialect handling. Future work willbetheuseofdeep learning modelsinmoresophisticatedways forintent recognition andfurtherexpandthelanguagesupported by the system. Personalized user interactions along with data-driven insights can also be added to improve the utility of the systemevenfurther.

ThisprojectmakesclearthetransformativepotentialofAIin educational contextsbyofferinga scalable, cost-effective solutionthatanyinstitutionworldwidecantakeup.Itsetsa benchmarkon chatbotsfortheapplicationsthatadvanced technologiescan beusedto meetthegrowingdemand for automatedandefficientqueryresolution.

REFERENCES

[1] Walaa Hassan, Shereen elBohy, Mina Rafik, Ahmed Ashraf, Sherif Gorgui, Michael Emil, Karim Ali. “An Interactive Chatbot for College Enquiry” Journal of ComputingandCommunicationVol.2,No.1,PP.20-28, 2023

[2] A.KousarNikhath,VijayaSaraswathiR,MDAbdulRab, NVenkataBharadwaja,LGouthamReddy,KSaicharan, CVenkatManasReddy "AnIntelligentCollegeEnquiry Bot using NLP and Deep Learning based techniques.” Janauary-2022

[3] Pavithra N, M A Bindhushree, Nanditha V, Pooja R, Shilpa. "AI-Based Automated Voice Responder for College Queries." International Journal of Innovative Research in Science, Engineering and Technology (IJIRSET)Volume13,Issue5,May2024

[4] A.Balamurugan,D.Thiruppathi,S.P.Santhoshkumar,K. Susithra. "Artificial Intelligence-Based Chatbot with Voice Assistance.” 2024 International Conference on TrendsinQuantumComputingandEmergingBusiness TechnologiesCHRIST(DeemedtobeUniversity),Pune LavasaCampus,India.Mar22-23,2024

[5] Sangeeta Kumari, Zaid Naikwadi, Akshay Akole, PurushottamDarshankar.“EnhancingCollegeChatBot Assistant with the Help of Richer Human Computer InteractionandSpeechRecognition.”Proceedingsofthe InternationalConferenceonElectronicsandSustainable CommunicationSystems(ICESC2020)IEEEXplorePart Number:CFP20V66-ART;ISBN:978-1-7281-4108-4

[6] Harshala Gawade, Prachi Vishe, Vedika Patil, Sonali Kolpe.“CollegeEnquiryChat-BotSystem.”International JournalofEngineeringResearch&Technology(IJERT) ISSN:2278-0181Vol.9Issue09,September-2020.

[7] YaseenE,SwamydossD.“CollegeChatbot.”International Research Journal of Modernization in Engineering TechnologyandSciencee-ISSN:2582-5208Volume:05 Issue:03,March-2023

[8] Tarun Lalwani, Shashank Bhalotia, Ashish Pal, Shreya Bisen, Vasundhara Rathod. “Implementation of a ChatbotSystemusingAIandNLP.”InternationalJournal of Innovative Research in Computer Science & Technology (IJIRCST) ISSN: 2347-5552, Volume-6, Issue-3,May-2018

[9] Amol Halvankar, Aakash Chaudhari, Aditi Wable, PriyankaBarkund,Dr.VishalPatil.“CollegeEnquiryFor Student using AI ChatBot.” International Scientific Journal of Engineering and Management ISSN: 25836129Volume:03Issue:04,April–2024

[10] Guruswami Hiremath, Aishwarya Hajare, Priyanka Bhosale,RasikaNanaware,Dr.K.S.Wagh.“Chatbotfor education system.” International Journal of Advance Research,IdeasandInnovationsinTechnology(IJARIIT) ISSN:2454-132XVolume4,Issue3,December-2020.