Inspiration
I wanted to create a language checking platform because many languages are marginalized and are hard to learn. AI models are also hard to develop for the average user.
What it does
How we built it
I used Markov chains as a quick AI language model that runs in the browser. I also allowed the user to create custom tokenizers and grammar rules with JavaScript. All of this is done entirely in the browser so that quick prototyping doesn't need a lot of complex setup.
Challenges we ran into
A big challenge that I ran into was just small bugs that needed to be fixed. They were annoying, but I got through them.
Accomplishments that we're proud of
I'm really proud that I was able to build this proof-of-concept in less than 2 days of work.
What we learned
I learned a lot about NLP and methods of validating grammar.
What's next for LinguaCheck
In the future, I plan to add more accurate and better AI language models through the use of technologies like WebGPU. I also plan to add more customization options like a custom syntax for manual grammar rules.
Log in or sign up for Devpost to join the conversation.