Inspiration

Although there are many flaws in the current system, it is undeniable that standardized tests are still a major factor in determining a student’s understanding of certain topics. Both of us are currently experiencing this period in our life: one is studying for the IGCSEs, while the other is preparing for AP exams. We’ve always attempted to find the best ways to study for these types of tests. After several years, we’ve identified that improving on aspects we consistently get incorrect is likely the best. However, doing this process manually is time-consuming and inefficient (said by many of our friends, too), and there are currently limited available platforms to do this process for us automatically. So, this became our idea.

What it does

Training Data

With natural language processing, our soon web application will be able to analyze users’ strengths and weaknesses on standardized tests. The AI model will be able to train with existing data from past year’s papers and cross-check answers with the mark scheme.

User Interface

The user’s written answer will be extracted and marked by the model with its corresponding mark scheme. Then, after a set of questions, the AI will come up with an analysis page for the user, identifying the strengths and weaknesses they have to work on. Users are able to select certain questions to view what they got wrong and a model answer for that question. This data is also stored on the user’s dashboard. After that, the AI will generate a practice exercise according to the things the users have to work on and master.

What's next for REVEAI_

With this idea, we believe that developing such a web application isn’t too far. The most challenging obstacle in developing would probably be gathering and making the data into a particular format. After we finish building the AI model, we will build the web application and create a more thorough business plan that is lucrative both for the users and us.

Built With

  • canva
  • capcut
  • figma
Share this project:

Updates