Natural Language Processing - Sentiment Analysis

Natural Language Processing

Sentiment Analysis

This notebook contains a sentiment analysis for this Kaggle challenge with an emphasis on making the NLP transparent. The “Behind the Code” sections show examples to create an understanding for the underlying transformations. This way, an intuition for the feature engineering process can be obtained.

I have used FastText pre-trained word-embeddings after I have pre-processed the text data with the nltk library.

You have to download the ‘wiki-news-300d-1M.vec’ if you like to run the full code.

The full project can be found here.