iServeSL is a centralised government-citizen interaction and smart web application for Sri Lankan citizens, enabling them to easily access and utilise the services offered by the government which will also integrate the distributed AI chatbot architecture in order to increase the overall user experience and provide seamless services. The aim of this application is to introduce an efficient, automated and intelligent approach of delivering government services online, with the help of AI assistance. Currently, many countries use web and mobile applications to offer their citizens with online government services. If we consider Sri Lanka, we find that most of the services are rendered manually where it is very time consuming, inefficient and sometimes inaccurate as well. Both the citizens and the government officials are irritated with the way government services are being provided. iServeSL is unique for its meticulous approach of addressing citizen inquiries using distributed chatbot architecture, which will provide its users with an AI assistance when they visit the web application. The cornerstone of iServeSL lies within this chatbot architecture. The main chatbot serves as a gateway directly engaging with the users. With the use of Natural Language Processing, it is able to filter and process user inquiries and route them to a chatbot specifically trained and designed to handle that particular service category. Each service offers its own dedicated chatbot that has been carefully trained and specialised to handle inquiries related to its field. These AI chatbots will also make use of Natural Language Generation technologies to generate the text or the answers back to the users in order to maintain an effective communication. The motivation to implement this application is that it ensures seamless access to government services from anywhere at any time. This distributed chatbot architecture improves the user experience by providing a superior customer service. These chatbots will handle all the user inquiries which are currently handled by different government workers and parties where it requires a great number of human resources and time to process. This dedicated chatbots will assure efficient and accurate problem solving since they are specialised in specific fields and trained to handle inquiries related to their relevant field. This approach guarantees faster responses and solutions, promoting a positive attitude towards the services. This approach will lead to a significant cost reduction in customer service and support as well as the government can allocate resources more efficiently afterwards. The researcher will use advanced technologies, utilising Google Dialogflow with Natural Language Processing, Natural Language Generation, and machine learning technologies with Python and TensorFlow. WebSocket API, React Chat UI, MongoDB, Ballerina and REST API will be used in order to make the platform more accessible and navigable.
Project Deliverables
Software Deliverables:
A centralized and interactive government-citizen web application implemented using distributed chatbot architecture, to enable easy access and utilise a wide range of government services for the citizens.
Documentation/Reports:
Project proposal
Contextual report
Thesis report
Reflective report
Test documentation
User manual
Artefact
After reviewing existing systems and research regarding G2C tech projects (Henman, P. 2020), (Starlin, E 2021), (de Sousa, W.G. et al 2019) it was concluded that G2C AI solutions are primarily focused on answering specific type of questions limited to a specific area using a virtual chatbot. Among them, ‘DubaiNow’ mobile application, developed and maintained by the Dubai government employs an AI chatbot called ‘Rashid’ with specific datasets where it can answer only questions on living and working in Dubai. iServeSL will expand this scope by introducing real-time AI assistance using distributed chatbot architecture which will enhance the effective communication between the user and the application.
Aim of the Project:
To implement a centralised, AI-driven web-based government-citizen interaction app with a real-time chat platform developed using distributed chatbot architecture for public service.
Objectives of the Project:
To identify and evaluate the current process of government services in Sri Lanka.
To identify existing technologies for improving public service with the AI-driven chatbot architecture.
To design the architecture of the distributed chatbots specialised for each service.
To implement the chatbot architecture with NLP and NLG techniques with machine learning models to support features of the application.
To implement a user-friendly web application including an interactive chat UI for service inquiries and assistance.
To secure the communication channels in the application.
To conduct tests and evaluate the application with real users.
Features of the Application:
Register new users/login existing users.
Enable citizens to report crimes.
Process all the incoming requests for police reports, validate them and generate police reports for the validated users.
Process all the incoming requests for character certificates and residence certificates issued by the village officer (government official known as ‘Grama Niladhari’ in Sri Lanka), validate them and generate the required certificates for the validated users.
Analyse the user inquiries sent via chat UI and provide responses or solutions by using dedicated chatbots in the system.
Note: This project will prioritise the incorporation of essential services (Police services and Village Officer services) that are widely used by citizens due to the time constraint the researcher is facing. However, it’s important to note that we can always develop additional features later on using the same methods and approaches.
The added value that the project provides:
Added value of the iServeSL lies within its ability to alter the government-citizen interactions, making the public services more accessible, efficient, and citizen centric. Through utilising advanced technologies and user-friendly interfaces, this application improves citizen engagement, streamlines administrative procedures, and fosters a more transparent and responsive government.
Intellectual Challenges:
Finding a large set of data regarding frequently asked questions and inquiries related to specific services.
Designing and implementation of distributed chatbot architecture using advanced NLP, NLG and machine learning models.
Learning and blending the new technologies within a tight timeline.
References:
de Sousa, W.G., de Melo, E.R.P., Bermejo, P.H.D.S., Farias, R.A.S. and Gomes, A.O., 2019. How and where is artificial intelligence in the public sector going? A literature review and research agenda. Government Information Quarterly, 36(4), p.101392. Available at: https://www.sciencedirect.com/science/article/abs/pii/S0740624X18303113.
The researcher will be using Agile development methodology to implement the iServeSL web application which comprises repetitive process of system design, implementation and testing. By following this methodology, the researcher will break the project up into smaller components where each component will be designed, developed, and tested within a time duration known as a ‘sprint’ in order to guarantee a smooth progression of the proposed system. Usually, one sprint lasts for two to four weeks of the month. Using Agile methodology will encourage continuous user testing and feedback in order to assure that this application will satisfy the needs of the citizens and the public officials.
During the requirement gathering phases, surveys will be conducted with the involvement of actual public officials and citizens. After gathering the required data sets, initial stage of developing the project will begin. Frequent prototyping and continuing improvements will support the project’s objectives to have a stable flow when developing the application using NLP, NLG, and AI technologies in order to successfully achieve the project aim in the end.
Relation Between the Project & the Degree Course [B.Sc. (Hons) Computer Science & Software Engineering]
Researcher will apply their perception of research methodologies to properly carry out a study on a topic and compose the required information.
During the system design phase, the researcher will identify the use cases of the proposed system and make use of Object-Oriented Analysis and Design concepts to accurately construct the architecture.
To maintain a clean and well-organized code base, the researcher will implement the proposed system by employing the Object-Oriented Programming concepts as well as Data Structures and Algorithms.
The researcher will utilise their expertise in Database Management Systems, Emerging Technologies and Frameworks to successfully address the intellectual challenges in the proposed system.
Resources
Technologies:
Ballerina programming language
Python
ReactJS
React Chat UI
MongoDB
TensorFlow
NLP: Natural Language Processing
NLG: Natural Language Generation
Google Dialogflow
REST API
WebSocket API
Tools:
Visual Studio Code
PyCharm by JetBrains
Botium
Asgardeo by WSO2
Hardware:
Personal computer
References
1. Books
- Raj, S., Raj, K. and Karkal, 2019. Building chatbots with Python. Apress.
2. Journal Papers
- Adamopoulou, E. and Moussiades, L., 2020. An overview of chatbot technology. In IFIP international conference on artificial intelligence applications and innovations (pp. 373-383). Springer, Cham. Available at: https://link.springer.com/chapter/10.1007/978-3-030-49186-4_31?ref=blog.min.io
- Adamopoulou, E. and Moussiades, L., 2020. Chatbots: History, technology, and applications. Machine Learning with Applications, 2, p.100006. Available at: https://www.sciencedirect.com/science/article/pii/S2666827020300062
- de Sousa, W.G., de Melo, E.R.P., Bermejo, P.H.D.S., Farias, R.A.S. and Gomes, A.O., 2019. How and where is artificial intelligence in the public sector going? A literature review and research agenda. Government Information Quarterly, 36(4), p.101392. Available at: https://www.sciencedirect.com/science/article/abs/pii/S0740624X18303113.
- Henman, P., 2020. Improving public services using artificial intelligence: possibilities, pitfalls, governance. Asia Pacific Journal of Public Administration, 42(4), pp.209-221. Available at: https://www.tandfonline.com/doi/abs/10.1080/23276665.2020.1816188
3. Web Sites with relevant information
- Starling, E. (2021) Smart Dubai’s Dubai now app introduces Ai Chatbot Rashid for: Course hero, Smart Dubai’s Dubai Now App Introduces AI Chatbot Rashid for | Course Hero. Available at: https://www.coursehero.com/file/116748958/Dubai-Now-artificial-intelligence-to-the-appdocx/[Accessed: 02 November 2023].
- Team (2022) Why G2C organizations should invest in Chatbot Development?, WS Group. Available at: https://www.webspiders.com/wp-blogs/why-g2c-organizations-should-invest-in-chatbot-development [Accessed: 01 November 2023].
Abstract
iServeSL is a centralised government-citizen interaction and smart web application for Sri Lankan citizens, enabling them to easily access and utilise the services offered by the government which will also integrate the distributed AI chatbot architecture in order to increase the overall user experience and provide seamless services. The aim of this application is to introduce an efficient, automated and intelligent approach of delivering government services online, with the help of AI assistance. Currently, many countries use web and mobile applications to offer their citizens with online government services. If we consider Sri Lanka, we find that most of the services are rendered manually where it is very time consuming, inefficient and sometimes inaccurate as well. Both the citizens and the government officials are irritated with the way government services are being provided. iServeSL is unique for its meticulous approach of addressing citizen inquiries using distributed chatbot architecture, which will provide its users with an AI assistance when they visit the web application. The cornerstone of iServeSL lies within this chatbot architecture. The main chatbot serves as a gateway directly engaging with the users. With the use of Natural Language Processing, it is able to filter and process user inquiries and route them to a chatbot specifically trained and designed to handle that particular service category. Each service offers its own dedicated chatbot that has been carefully trained and specialised to handle inquiries related to its field. These AI chatbots will also make use of Natural Language Generation technologies to generate the text or the answers back to the users in order to maintain an effective communication. The motivation to implement this application is that it ensures seamless access to government services from anywhere at any time. This distributed chatbot architecture improves the user experience by providing a superior customer service. These chatbots will handle all the user inquiries which are currently handled by different government workers and parties where it requires a great number of human resources and time to process. This dedicated chatbots will assure efficient and accurate problem solving since they are specialised in specific fields and trained to handle inquiries related to their relevant field. This approach guarantees faster responses and solutions, promoting a positive attitude towards the services. This approach will lead to a significant cost reduction in customer service and support as well as the government can allocate resources more efficiently afterwards. The researcher will use advanced technologies, utilising Google Dialogflow with Natural Language Processing, Natural Language Generation, and machine learning technologies with Python and TensorFlow. WebSocket API, React Chat UI, MongoDB, Ballerina and REST API will be used in order to make the platform more accessible and navigable.
Project Deliverables
Software Deliverables:
Documentation/Reports:
Artefact
After reviewing existing systems and research regarding G2C tech projects (Henman, P. 2020), (Starlin, E 2021), (de Sousa, W.G. et al 2019) it was concluded that G2C AI solutions are primarily focused on answering specific type of questions limited to a specific area using a virtual chatbot. Among them, ‘DubaiNow’ mobile application, developed and maintained by the Dubai government employs an AI chatbot called ‘Rashid’ with specific datasets where it can answer only questions on living and working in Dubai. iServeSL will expand this scope by introducing real-time AI assistance using distributed chatbot architecture which will enhance the effective communication between the user and the application.
Aim of the Project:
To implement a centralised, AI-driven web-based government-citizen interaction app with a real-time chat platform developed using distributed chatbot architecture for public service.
Objectives of the Project:
Features of the Application:
Note: This project will prioritise the incorporation of essential services (Police services and Village Officer services) that are widely used by citizens due to the time constraint the researcher is facing. However, it’s important to note that we can always develop additional features later on using the same methods and approaches.
The added value that the project provides:
Added value of the iServeSL lies within its ability to alter the government-citizen interactions, making the public services more accessible, efficient, and citizen centric. Through utilising advanced technologies and user-friendly interfaces, this application improves citizen engagement, streamlines administrative procedures, and fosters a more transparent and responsive government.
Intellectual Challenges:
References:
de Sousa, W.G., de Melo, E.R.P., Bermejo, P.H.D.S., Farias, R.A.S. and Gomes, A.O., 2019. How and where is artificial intelligence in the public sector going? A literature review and research agenda. Government Information Quarterly, 36(4), p.101392. Available at: https://www.sciencedirect.com/science/article/abs/pii/S0740624X18303113.
Henman, P., 2020. Improving public services using artificial intelligence: possibilities, pitfalls, governance. Asia Pacific Journal of Public Administration, 42(4), pp.209-221. Available at: https://www.tandfonline.com/doi/abs/10.1080/23276665.2020.1816188
Starling, E. (2021) Smart Dubai’s Dubai now app introduces Ai Chatbot Rashid for: Course hero, Smart Dubai’s Dubai Now App Introduces AI Chatbot Rashid for | Course Hero. Available at: https://www.coursehero.com/file/116748958/Dubai-Now-artificial-intelligence-to-the-appdocx/ [Accessed: 02 November 2023].
Methodology
The researcher will be using Agile development methodology to implement the iServeSL web application which comprises repetitive process of system design, implementation and testing. By following this methodology, the researcher will break the project up into smaller components where each component will be designed, developed, and tested within a time duration known as a ‘sprint’ in order to guarantee a smooth progression of the proposed system. Usually, one sprint lasts for two to four weeks of the month. Using Agile methodology will encourage continuous user testing and feedback in order to assure that this application will satisfy the needs of the citizens and the public officials.
During the requirement gathering phases, surveys will be conducted with the involvement of actual public officials and citizens. After gathering the required data sets, initial stage of developing the project will begin. Frequent prototyping and continuing improvements will support the project’s objectives to have a stable flow when developing the application using NLP, NLG, and AI technologies in order to successfully achieve the project aim in the end.
Relation Between the Project & the Degree Course [B.Sc. (Hons) Computer Science & Software Engineering]
Researcher will apply their perception of research methodologies to properly carry out a study on a topic and compose the required information.
During the system design phase, the researcher will identify the use cases of the proposed system and make use of Object-Oriented Analysis and Design concepts to accurately construct the architecture.
To maintain a clean and well-organized code base, the researcher will implement the proposed system by employing the Object-Oriented Programming concepts as well as Data Structures and Algorithms.
The researcher will utilise their expertise in Database Management Systems, Emerging Technologies and Frameworks to successfully address the intellectual challenges in the proposed system.
Resources
Technologies:
Tools:
Hardware:
References
1. Books
2. Journal Papers
3. Web Sites with relevant information