SchemeBot - AI Enabled Chatbot for Government Schemes

Abstract

Government assistance programs are essential to socioeconomic growth, but their influence is frequently constrained by a lack of accessibility and public awareness. We introduce SchemeBot, an intelligent chatbot created to close this information gap by giving citizens quick access to details on a variety of government initiatives. With a focus on three core areas—agriculture, healthcare, and women and children—SchemeBot uses a solid backend foundation in conjunction with an organized frontend interface to provide a smooth user experience. The chatbot uses sophisticated text preprocessing methods, such as tokenization, bag-of-words, case normalization, and lemmatization, to improve the accuracy and clarity of data using JSON-formatted input. A multi-layer neural network with ReLU activation functions powers the chatbot\'s response creation, ensuring accurate and contextually relevant interactions. The system architecture, development process, and assessment metrics of SchemeBot are described in this study, which also highlights its potential as a scalable solution to improve the efficacy and outreach of government projects. According to our research, SchemeBot greatly enhances information accessibility and is an essential instrument for raising public awareness of welfare programs.

Introduction

I. INTRODUCTION

Fundamental to socio-economic development are government initiatives and welfare programs that aim to improve communities, assist marginalized people, and deal with urgent social challenges. Even with their importance, these programs' accessibility and awareness are frequently restricted, particularly in rural and impoverished areas. Because of disjointed communication channels and the difficulty of navigating several government websites, citizens frequently have difficulties in accessing pertinent information. Many recipients are unable to take advantage of the chances meant to enhance their quality of life because of this information gap.

In response to this difficulty, we created SchemeBot, an intelligent chatbot that offers a convenient interface for obtaining details on different government initiatives. SchemeBot concentrates particularly on three important areas: healthcare, farming, and women and children. SchemeBot uses natural language processing (NLP) and machine learning to make knowledge retrieval easier. Users can communicate with a virtual assistant that provides accurate and contextually relevant answers.

The chatbot's backend is supported by a strong framework that can handle a variety of user questions, while its frontend is made for an easy-to-use interface. To improve consistency and interpretation, the JSON-formatted training data for SchemeBot is thoroughly preprocessed using techniques including tokenization, bag-of-words transformation, case normalization, and lemmatization. The underlying neural network is optimized to produce precise replies that are suited to the user's requirements. It is constructed using three layers and ReLU activation functions.

This paper describes SchemeBot's architecture, implementation, and evaluation in detail, emphasizing how it might completely change the way the public interacts with public data. SchemeBot seeks to revolutionize public welfare communication by utilizing AI and machine learning approaches to make it more user-centric, efficient, and accessible. In order to show how this technology may be used as a scalable tool to improve program outreach and citizen involvement, our study examines how it bridges the gap between public activities and government projects.

II. LITERATURE REVIEW

Recent years have witnessed a notable increase in the use of chatbots and artificial intelligence (AI) in the public sector, mostly because of the potential benefits they offer for enhancing citizen engagement and service delivery. Chatbots are a potent tool for delivering quick and easily accessible information because they use natural language processing (NLP) techniques to comprehend and reply to user queries. This review of the literature looks at recent studies and technological advancements in the area of artificial intelligence (AI)-based chatbots, with an emphasis on how these chatbots might be used to spread knowledge about public welfare initiatives and government schemes.

Dr. G. Nanthakumar[1]The design of a chatbot that uses natural language processing and machine learning to help citizens with government programs is explained in this paper. From a JSON database, the system retrieves pertinent answers to queries by classifying them using natural language processing (NLP) and machine learning. Features like sentiment analysis, handling of several turns of interaction, and extended reach will improve the chatbot's functionality. By using chatbots to perform public services, the adoption of these systems will increase inclusivity, efficiency, and transparency and empower citizens to take an active role in government projects and make informed decisions.

Indhumathi[2] A Natural Language Processing (NLP) chatbot has been created to help citizens ask questions about their eligibility for various government initiatives using an interactive interface. The AI-powered chatbot generates, understands, and interprets human conversation using sophisticated algorithms and natural language processing skills. It is intended to close the gap between public activities and government actions by providing real-time responses to user requests. Future development will involve expanding the range of schemes covered by the chatbot, improving its natural language processing skills, and incorporating it into an Android application. The chatbot facilitates individualized encounters, which in turn encourages transparency and well-informed decision-making.

Mrs. K. Sowndharya[3] The Government Scheme Navigator Chatbot is an approachable platform that informs citizens about government schemes through the use of artificial intelligence and natural language processing. It makes it easier for citizens to get pertinent information by streamlining the navigation of intricate websites or lengthy documentation. The chatbot gives personalized answers depending on each user's situation and provides comprehensive information on contacts, application procedures, benefits, and eligibility requirements. It complies with data protection laws and keeps an extensive database of government programs. In order to enhance its responses over time, the chatbot also makes use of machine learning methods. The platform eliminates conventional obstacles to accessibility and promotes better involvement and understanding of government services. Ensuring a smooth user experience, the Government Scheme Navigator Chatbot is a cutting-edge platform that encourages diversity and transparency.

Mr. Prakash. V[4] Researchers are utilizing a variety of soft-computing techniques to improve the user-friendliness of chatbots, which are an effective tool for automatically and accurately responding to customer inquiries. With the use of cutting-edge NLP algorithms and machine learning models, our initiative, Scheme Setu, seeks to empower individuals by offering a complete data hub for government-sponsored loans and insurance plans.

The eight-stage technique ensures the effectiveness and dependability of the chatbot through data sourcing, model integration, and user-friendly interaction. Large datasets must be stored in a chat program for the system to work. TensorFlow is used to train an RNN model, and checkpoints are saved every 5,000 steps.

Shomitro Kumar Ghosh[5] The integration of chatbots into digital environments is growing; they act as digital assistants and improve e-government systems by offering a range of e-services. The use of AI in governance is advancing in Bangladesh, despite regional differences in technology readiness and application. Private sector collaborations are crucial in this regard. Essential functions, including publishing, calling, and defining services, are supported by the e-services architecture, which encourages interoperability and just-in-time integration. Notwithstanding the potential advantages, e-government has difficulties in identifying and putting into practice appropriate software solutions; therefore, the creation of widely recognized e-service standards is required to guarantee broad acceptance.

SHREYAS S [6] Through the consolidation of data on government-sponsored loans and insurance plans, Scheme Setu is a chatbot designed to improve user experience and financial literacy. To deliver precise and pertinent answers to intricate customer inquiries, it makes use of cutting-edge machine learning models and Natural Language Processing (NLP) algorithms. The Mega Bot seeks to empower people financially and break down barriers by democratizing possibilities and societal strata. Data sourcing, model integration, Langchain connectivity, data source integration, user-friendly interface, testing and validation, documentation, and reporting are the eight phases of the research approach.

The caching mechanism of the system guarantees quick retrieval of frequently accessed data, and LLMs enhance their intelligence to offer a customized and effective user experience. The online-based interface provides a fluid user experience across various browsers and devices, while the Android application is tailored for the Android operating system. Scheme Setu is committed to strengthening its role as an empowerment catalyst while striving to close the gap between citizens and government-sponsored financial aid.

III. METHODOLOGY

A. Backend Development

1) Django framework

1. Django, a high-level Python web framework renowned for its scalability, clean code design, and quick development capabilities, is used to create the chatbot system's backend. Django's built-in utilities, such as manage.py for operating the server and executing database migrations, make project management simple.
2. Organization of Files: Important files are db.sqlite3, which specifies that SQLite is being used as the database, and manage.py, which manages server operations.
3. Object-Relational Mapping (ORM): By interacting with the SQLite database through Django's ORM, tasks involving data retrieval and modification are streamlined without the need for raw SQL queries.

2) Python Programming

Python is used to create the basic logic of the chatbot, which includes the several domain-specific bots like educationbot.py, farmerbot.py, healthbot.py, and womenbot.py. This approach makes use of Python's ease of handling logical operations and its compatibility with Natural Language Processing (NLP) technologies.

3) SQLite Database

The backend database used to hold user information, chatbot replies, and conversation logs is called SQLite. For lightweight, server-less applications that value mobility and simplicity throughout development, its file-based design is perfect.

B. Chatbot Development

1) Natural Language Processing

1. NLP techniques are used by the chatbot to comprehend user input, determine user intents, and provide pertinent responses. The intentions.json file contains pre-defined intents that are used by a pattern-matching technique to drive intent recognition.
2. Files and Data Handling

• The primary training set is contained in intents.json, which describes the goals, patterns, and potential replies of the chatbot.
• The education.json, farmer.json, health.json, and women.json domain-specific JSON files offer customized answers to user inquiries, improving the relevancy and precision of the chatbot.

2) Modular Design

1. By breaking out each domain into distinct Python scripts (educationbot.py, farmerbot.py, etc.), the chatbot has a modular architecture that makes it simpler to maintain and update individual parts without affecting the system as a whole.
2. Benefits of Modularity: By facilitating the easy addition of new domains, this strategy improves scalability and maintainability while strengthening the chatbot's capacity to respond to a wide range of user requests.

C. Frontend Development

1) HTML, CSS, and JavaScript:

1. The chatbot's user interface (UI) is created with JavaScript for interactivity, CSS for styling, and HTML for structure.
2. Frameworks and Libraries:

• Using Bootstrap, responsive user interface components may be designed so that the chatbot's layout can adjust to various screen sizes and devices.
• jQuery: Improves the chatbot's interaction by streamlining JavaScript code and enabling dynamic content changes.
• Smooth.js: Integrated to enhance the way information is presented in the chatbot user interface by generating responsive carousels.

D. Static Assets Management

Static assets, such as photos in the.jpg and.png formats, are used in the project to improve the web interface's visual style, including banners and logos. To make the experience more interesting and user-friendly, several components are included.

E. Methodologies and Development Practices

1. Separation of Concerns: By dividing chatbot functions among domain-specific modules, the modular architecture adheres to the separation of concerns concept, improves code structure, and facilitates troubleshooting.
2. Intent-Driven Design with JSON: JSON-based datasets that specify potential user inputs (patterns) and related bot responses are the foundation for the chatbot's intent recognition. This method makes it possible to easily update and expand the bot's knowledge base.
3. Data-Driven Reaction Mechanism: In response to customer inquiries, the chatbot dynamically retrieves answers from JSON databases. It can deliver accurate and contextually relevant responses because of its data-driven approach, especially in specialized industries like education and healthcare.
4. The full-stack development methodology: The system uses a full-stack development methodology, with HTML, CSS, and JavaScript combined for the frontend and Django for the backend. The well-integration of the logic processing and user interface components is guaranteed by this all-inclusive development methodology.
5. Architecture that is RESTful: RESTful APIs are used by Django's backend architecture to facilitate effective communication between the logic of the chatbot and the frontend interface, encouraging a uniform method of managing data and producing responses.
6. Iterative or agile development: Agile approaches were used throughout the development process, which allowed the chatbot to be developed gradually with ongoing testing, deployment, and functional improvement. Modular improvements were made possible by this iterative process, which guaranteed that every bot module could be created and improved on its own.

F. Modular Bot Design

The project uses a modular bot architecture, with each chatbot module concentrating on a particular industry, such as healthcare, agriculture, or education. This division of responsibilities permits:

1. Maintenance Ease: Each bot module can be updated and maintained separately from the others without impairing their ability to perform.
2. Scalability: The chatbot's ability to cover more areas of interest to users can be improved by adding new modules to address different domains.

IV. RESULT

In order to create a more scalable and maintained chatbot, the project successfully included a modular design by dividing chatbot functionality into domain-specific components. Because each module had a clear function and was well-defined, the development process was expedited and accurate replies were produced. With intents specified in JSON files, the chatbot leveraged effective Natural Language Processing (NLP) techniques to comprehend customer inquiries and provide relevant responses. The entity recognition and intent-driven design methodologies were successfully applied. Backend functions were managed by the Django backend, while lightweight data storage was provided by the SQLite database. The frontend's responsive and user-friendly interface was made possible by the usage of HTML, CSS, JavaScript, and Bootstrap. Because the chatbot's responses were constantly generated from JSON information, updating and growing the bot's knowledge base was made simple and flexible. The chatbot's end-user experience was enhanced by this data-driven response strategy.

Figure 1: User Interface

Figure 2: Chatbot UI for Women’s Schemes

Conclusion

The chatbot framework is a powerful full-Stack solution that combines a JSON-based data structure and Django with NLP features. Because of its modular design pattern, adding or updating features is simple and doesn\'t interfere with the system\'s current functionality. Future advancements are accommodated by the scalable and maintained nature of the chatbot. The backend\'s RESTful architecture makes sure that frontend and backend components communicate with each other seamlessly. Continuous testing, deployment, and improvement are made possible by the agile development technique, which guarantees that the product can quickly adjust to changing user needs.

References

[1] Nanthakumar, G., Pazhanivel, K., Saivasanth R, Sri Hari R M, & Vignesh R. (2024). Government Schemes Chatbot Using Machine Learning. In International Journal of Advanced Research in Science, Communication and Technology (Vol. 4, Issue 6) [Journal-article]. https://www.ijarsct.co.in [2] Indhumathi, Balaji, Elavarasan, Navaneethakrishnan, Nivas, & Computer Science And Engineering, Nandha College Of Technology, Erode, India. (2024, April). AI BASED CHATBOT FOR GOVERNMENT SCHEME ELIGIBILITY [Press release]. https://www.doi.org/10.56726/IRJMETS54131 [3] Sowndharya, N. M. K., M, N. M., M, N. S. S., & J, N. S. (2024). Government scheme navigator Chatbot. International Journal of Advanced Research in Science Communication and Technology, 444–448. https://doi.org/10.48175/ijarsct-17572 [4] Prakash, V., Surya, M., Shri Mahavishnu, N., Gobinath, T., & Chettinad College of Engineering and Technology. (2024). A survey of Chatbot for providing government schemes to citizens. International Journal of All Research Education and Scientific Methods (IJARESM), 685. https://www.ijaresm.com [5] Ghosh, S. K., Daffodil International University, Bhuiyan, M. T., Daffodil International University, Chowdhury, R., Daffodil International University, Jabiullah, Md. I., & Daffodil International University. (2020). A chatbot-based e-Services for e-Government. In Conference Paper. https://www.researchgate.net/publication/342747448 [6] S, S., TIWARI, T., V, T., VIKRAM, & Department of ISE, Presidency University, Bangalore, India. (2024). Scheme Setu: a chatbot for government schemes. In IRE Journals (Vol. 7, Issue 7, pp. 119–120).

Copyright

Copyright © 2024 Manisha Mali, Akshada Mane, Sneha Rahate, Vaishnavi Shinde, Akshata Vibhute, Garv Chopda. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.