spellit

• 

  sample

Inspiration

I learnt about TRIE in a data structures class and thought it can be used in an auto-correction tool. On some research, I found that I clearly wasn't the first one to think of that. I coded a TRIE in C++ last week and decided to build one in JS for this hackathon.

What it does

It creates a TRIE using a 3000 word database and waits for you to enter a word.

How we built it

In most of the languages, around 90% of conversations are made of a small set of words. In English, most common 3000 words can cover up to 95% of our conversations. Using this property, I scrapped a list of 3000 most common words and created a word recommender out of it.

What we learned

Integration of various technologies was pretty cool.

What's next for spellit

Spellit will soon be launched as a chrome extension. It can be made more inclusive by adding more languages and words in the database.

Built With

• astradb
• css
• html
• javascript
• python