Smart Scheme Recommender

Smart Scheme Recommender for Every Citizen is a government scheme recommendation system designed to simplify the process of discovering and applying for welfare schemes

Comment

Inspiration

Many citizens are unaware of the welfare schemes they are eligible for due to scattered information and lack of accessibility. We wanted to solve this gap by creating a system that makes government schemes more approachable, especially for people in rural areas and non-English speakers.

What it does

The Smart Scheme Recommender is a web application that recommends personalized government welfare schemes based on a user’s attributes such as age, income, occupation, category, and physical condition. It also supports Tamil and Hindi languages, provides summarized eligibility details, and even includes text-to-speech features for accessibility.

How we built it

Frontend: HTML5, CSS, JavaScript, Bootstrap

Backend: Python (Flask) + MySQL for storing scheme data

Machine Learning: Basic ML algorithms and NLP models to match user profiles with eligible schemes

Additional Features: Multilingual support and dual-mode summarization for concise scheme details

Challenges we ran into

Collecting and structuring authentic government scheme data was difficult.

Implementing multilingual NLP and ensuring accuracy in translations was challenging.

Balancing performance and accessibility features (like text-to-speech) without compromising usability.

Accomplishments that we're proud of

Built a functional end-to-end prototype that recommends schemes with good accuracy.

Integrated Tamil and Hindi support to make the platform inclusive.

Presented the project at a state-level hackathon (NIRAL Thiruvizha) and later published a research paper in IEEE Xplore.

What we learned

How to apply machine learning and NLP models to solve real-world problems.

The importance of UI/UX in accessibility, especially for government platforms.

Effective team collaboration using GitHub for version control and task management.

What's next for Smart Scheme Recommender

Expand the database to include all state and central government schemes.

Improve ML algorithms for more accurate recommendations.

Add mobile app support for wider accessibility.

Collaborate with government bodies for real-time updates of scheme data.

Share this project:

Updates