Using word2vec, we can graph and see our path to self improvement
There are a lot of good guides and tools for exploring word2vec, I'd recommend taking a look at all of them :
https://github.com/dominiek/word2vec-explorer
https://github.com/pg2455/U.S-Presidential-Speeches
https://matplotlib.org/index.html
Feel free to use the Jupyter Notebook version.
"word2vec-self-improvement.py" is a freestanding Python Application that does the same task. To run it, type
word2vec-self-improvement.py fun carefree tasty wise older loving
Where the first 3 words indicate your current adjectives and the last indicate your goal state.
If you are running it with qsub, (like on the Intel DevCloud), you can just submit the launch file to the que by
qsub launchimprove
This will also download the pre-trained models.
Emjoy!