AlgoArena — From Solving Problems to Understanding Them
Introduction
Most coding platforms focus on one question:
Did your code pass?
But passing test cases does not always mean that a learner truly understands the algorithm behind the solution.
AlgoArena is a learning-focused coding platform designed to bridge that gap. It combines coding challenges with AI-powered understanding evaluation to help learners move beyond simply getting the correct output.
The goal is simple:
Solve → Explain → Understand → Improve
The Problem
When practicing competitive programming or DSA, learners usually receive feedback based on whether their code passes the test cases.
However, this doesn't tell us:
- Does the learner understand why the algorithm works?
- Do they understand its time and space complexity?
- Can they explain the key concept behind their solution?
- Are they memorizing solutions instead of understanding them?
- Which concepts do they need to improve?
Traditional coding judges are excellent at checking correctness, but they don't necessarily measure conceptual understanding.
AlgoArena was built to address this gap.
How AlgoArena Works
AlgoArena introduces an additional Understanding Check after a successful coding submission.
1. Solve the Coding Problem
The learner selects a DSA problem and writes a solution.
The submission is evaluated against test cases to determine whether the solution is correct.
2. Understanding Check
If the code is accepted, AlgoArena generates conceptual questions related to the problem.
Instead of immediately showing whether each answer is good or bad, the learner is asked to explain their reasoning in their own words.
For example, a problem involving a HashMap may ask:
- Why is a HashMap useful for this problem?
- What information should be stored in it?
- What is the time complexity?
- Why must the current element be checked before being inserted?
This turns the coding exercise into a deeper learning experience.
3. AI-Powered Evaluation
The submitted explanations are evaluated using Google's Gemini API.
The evaluation considers factors such as:
- Understanding level
- Relevant concept
- Quality of explanation
- Conceptual gaps
The system categorizes understanding into levels such as:
GOOD, PARTIAL, and POOR.
The individual evaluation is handled privately so that the learner can focus on completing the entire understanding check.
4. Final Feedback
After completing the understanding questions, AlgoArena generates a final feedback summary.
The learner can see a structured overview of their understanding rather than just a simple "Accepted" result.
This helps answer the more important question:
"Do I actually understand the solution I just wrote?"
Personalized Learning Insights
AlgoArena also keeps track of understanding across problems.
The backend can generate information such as:
- Understanding summaries
- Weak concepts
- Concept history
- Concept progress
- Concept status
- Recommended problems
- Overall learning progress
This allows the platform to move from simply judging code to tracking the learner's conceptual development.
For example, if a learner repeatedly struggles with a particular concept, AlgoArena can identify that concept and recommend additional problems related to it.
Technology Stack
Frontend
- React
- Vite
- React Router
- CSS
Backend
- Java
- Spring Boot
- Spring Data JPA
- REST APIs
- MySQL
AI
- Google Gemini API
Top comments (0)