spark_ml

The purpose Of this project is to demonstrate the usage of spark mllib from training to predicting ecommercial product's category.

Training Stage:

1. Uses tfidf to vectorize text context
2. Uses Cliffisier Algorithms such as Naive Bayes, Ovr Logistic Regression, Random Forest to train the model

Predict Stage:

1. Load models from hdfs
2. Use models to classify products' category