My solution for Kaggle 'Home Depot Product Search Relevance' competition

This is my solution for the 'Home Depot Product Search Relevance' competition on Kaggle, which scored 0.45766 (76/2125) on Public Leaderboard and 0.45751 (83/2125) on Private Leaderboard.

Kaggle ID : http://www.kaggle.com/ssjatmhmy

Generating the solution

step 1
Run the following command to generate word2vec training data 'data.text'.
cd code
python3 gen_text.py
Then copy the generated file 'data.text' into the 'tools-w2v' folder.

step 2
Run the following command to generate word2vec model, and then re-enter the 'code' folder:
cd tools-w2v
python3 train_word2vec_model.py data.text hd
cd ../code

step 3
Run the following command to generate cleaned data:
python3 run.py -PrepareData

step 4
Run the following command to generate tsne features. Note that it has to be run with python2, because the lib we used does not support python3.
python2 python2_tsne.py

step 5
Run the following command to generate all features:
python3 run.py -GenerateFeatures

step 6
Run the following command to generate submission:
python3 run.py -GetSubmission