My bud Micah and I were working on a coding project. Micah raps a lot and produces music, so while we were staying up late he kinda mumbled lyrics and we referenced rap songs a lot. I tried freestyling myself during my trips to and from home, but I always had to pace myself and I couldn't think of rhymes in a reasonable time. So I suggested to Micah that we might work on something like that, and started writing it in python. We met Nishanth and Rahul and they loved the idea and turned out to be node geniuses so we decided to team up and hammer it out.
What it does
It's simple! It takes a line you say and gives you words to rhyme with it making for a sleek creative experience.
How WE built it
Haha. We had it pretty well planned out and divided up the roles and helped each other. Nishanth was the best at web stuff, but helped us and distributed tasks so that we were always working on something.
Challenges WE ran into
We couldn't get rhymes fast enough! The original dictionary that we used contained 130k words, and to generate a quick algorithm we had to find a way to cope with the slow speed. We tried making a gigantic hash table, but using the libraries we had, it would've taken 18 hours on our machines, and we didn't want to experiment with multi threading the code because we weren't sure we had enough time to experiment with distribution. We solved it by filtering out words by common use frequency.
Accomplishments that I'm proud of
WE FINISHED IT. WE WORKED TOGETHER. EVERYONE CONTRIBUTED. TEAMMATES ARE NICE, TALENTED, AND HARD WORKING!
What I learned
What's next for Spitfire
Data Mining Improving word prediction Improving sentence ending