Sentiment analysis has been something I've been interested in for a long time. From my first python script pulling reddit comments to this program now, I've found it compelling and challenging, while also having a low barrier of entry (the first iterations of this compared each word to known "Bad" or "Good" words). I like making all of my own components for this as well. I could easily use vectors or map for much of this, but I like understanding and specializing my own data structures. The project will be split up into two major components: vectorization and a trained neural net to determine semantic meaning.
- Finish b_tree
- Finish parsing into windows
- Make word window manager class
- Make graph for words
- Finish testing
- Make matrix of word associations