Korean Movie Reviews and Historical Ratings of the Reviewer

Published: 24 May 2022| Version 6 | DOI: 10.17632/jb5knzh8yv.6


The data sets are Korean movie reviews focusing on historical textual reviews and ratings of the reviewer. We collect them from Naver Movie, which has the largest movie database in South Korea. It consists of two kinds of data sets: 1) validation and 2) learning. In the validation data sets, the total number of reviews is 96,392 is included for 7 movies. In the learning data sets, the total number of reviews for training is 30,720 and that for testing is 7,680 for 6 movies. The attributes of each data include the review ID, user ID, review, rating, the number of helpfulness votes, the number of unhelpfulness votes, and timestamp. Especially, the number of helpfulness votes and that of unhelpfulness votes of each movie review, which had been evaluated by the other reviewers, can be used to judge the helpfulness of the review. Moreover, we collect all the historical and textual reviews and ratings of the reviewer who writes the review for the target movie. The historical ratings of the reviewer can be crucial evidence to judge the characteristics of the movie reviewer.



Seoul National University of Science and Technology


Data Science, Big Data Analytics, Data Engineering