Predict the average review ratings of products on Amazon
Online product reviews provided by consumers who previously purchased products have become a major information source for consumers and marketers regarding product quality. Research has shown that consumer online product ratings reflect both the customers' experience with the product and the influence of others' ratings. Websites prominently display consumers' product ratings, which influence consumers' buying decisions and willingness to pay.
There are 16 independent variables and 1 dependent variable. There are nearly 10k observations combining both train and test dataset.
- uniq_id
- product_name
- manufacturer
- price
- number_available_in_stock
- number_of_reviews
- number_of_answered_questions
- average_review_rating
- amazon_category_and_sub_category
- customers_who_bought_this_item_also_bought
- description
- product_information
- product_description
- items_customers_buy_after_viewing_this_item
- customer_questions_and_answers
- customer_reviews
- sellers
The test set contains all the above variables except for average_review_rating, which is to be predicted.
This dataset was created by PromptCloud's in-house web-crawling service.