Goodreads Book Reviews (Popular and Non-Popular)

The review corpus used here consists of a collection of Goodreads book reviews obtained from the Kaggle website. Originally, it consists of around 10,000 reviews written towards top 100 science fiction books (ranked based on Goodreads ratings). However, upon examination, we find that a significant number of reviews comprise various types of issues, including missing values for the review text and like count, as well as duplicate sentences. We remove duplicate sentences from the reviews and exclude problematic reviews with missing fields. After clean-up, the corpus consists of 2259 popular and 2555 non-popular reviews.