product_sentiment_machine_hack

Classify the sentiment (4-way classification) of user reviews of products based on the review text and product type (e.g. Tablet, Mobile, etc.). Intuitively, we expect most of the predictive signal to lie in the text, but predictions can be further improved by accounting for the fact that certain types of products tend to receive certain user sentiment. Representing a relatively simple multimodal task with only a single text feature and one categorical feature, this dataset originally stems from a 2020 MachineHack prediction competition: https://machinehack.com/hackathons/product_ sentiment_classification_weekend_hackathon_19/overview Dataset found from the paper: Benchmarking multimodal automl for tabular data with text fields. arXiv preprint arXiv:2111.02705.