ChenglongChen/kaggle-HomeDepot
Third-place Kaggle solution that predicts search relevance scores for e-commerce product queries using NLP and ensemble ML techniques.
Not currently ranked — collecting fresh signals.
star history
This repository contains the Turing Test team’s solution for the Home Depot Product Search Relevance competition. The approach uses natural language processing with semantic matching techniques to predict how relevant search results are to user queries. It combines multiple models including word embedding approaches via gensim, neural networks built with Keras, and gradient boosting via xgboost. The solution features extensive feature engineering including text preprocessing, relevance scoring, and ensemble methods to achieve top competition rankings.
Frequently asked
- What is ChenglongChen/kaggle-HomeDepot?
- Third-place Kaggle solution that predicts search relevance scores for e-commerce product queries using NLP and ensemble ML techniques.
- Is kaggle-HomeDepot open source?
- Yes — ChenglongChen/kaggle-HomeDepot is open source, released under the MIT license.
- What language is kaggle-HomeDepot written in?
- ChenglongChen/kaggle-HomeDepot is primarily written in Python.
- How popular is kaggle-HomeDepot?
- ChenglongChen/kaggle-HomeDepot has 468 stars on GitHub.
- Where can I find kaggle-HomeDepot?
- ChenglongChen/kaggle-HomeDepot is on GitHub at https://github.com/ChenglongChen/kaggle-HomeDepot.