This is my submission for the Texata 2015 finals (http://texata.com/), as one of the 12 contestants that made it to this round. With this submission I managed to get the second prize in the competition!

The main file to read to understand my submission is the report python notebook. References to the purpose of the rest of files is also provided there.

Also as a part of the challenge all the contestants were required to prepare a 5 minutes pitch as a summary of the work performed. Under the presentation folder you can find my pitch slides.

All the contents of the repository are exactly those as the ones submitted for the actual competition, with the exception of this README.md file, the LICENSE file, and the presentation slides, which got a significant revamp after I was requested to present it as part of the Texata 2015 summit.

Note that since some datasets used here are either CISCO property or too large to upload to Git, these are not included in the repository. Similarly, trained models were too large to upload and so are not included. Still you can download the Stack Exchange dumps and the english Wikipedia polyglot models and give it a try!

• Polyglot pretrained embeddings: https://sites.google.com/site/rmyeid/projects/polyglot
• Stack Exchange dumps: https://archive.org/details/stackexchange