Amazon fine food review - #Sentiment analysis
The analysis is to study #Amazon food review from customers, and try to predict whether a review is positive or negative. The dataset contains more than 500K reviews with number of upvotes & total votes to those comments.
Table of Content
Data preparation
Score prediction
Logistic regression model on word count
Logistic regression model on TFIDF
Logistic regression model on TFIDF + ngram
Upvote prediction
Data preview
Resampling due to imbalanced data
Logistic regression model on word count
Logistic regression model on #TFIDF + #ngram
Study on non-context features
In-depth study on user behaviour
Thumbnail image credits goes to: https://www.thepythoncode.com/media/a...
コメント