MovieLens-32M扩展数据集
收藏资源简介:
MovieLens-32M扩展数据集是由滑铁卢大学的研究团队创建的,旨在为推荐系统提供新的评估目标。该数据集包含了51位MovieLens用户的评分档案和31236条关于未评分电影的相关性判断。这些判断是通过用户对推荐电影的海报、标题、年份、剧情摘要等信息的评估得出的,目的是预测用户可能会感兴趣观看的电影。该数据集的应用领域是推荐系统,特别是用于评估推荐算法的性能,减少或消除评估中的流行度偏差。
The MovieLens-32M extended dataset was created by a research team from the University of Waterloo, aiming to provide novel evaluation benchmarks for recommender systems. This dataset contains rating profiles of 51 MovieLens users and 31,236 relevance judgments for unrated movies. These judgments are derived from users' evaluations of information such as posters, titles, release years, and plot summaries of recommended movies, with the goal of predicting movies that users may be interested in watching. This dataset is applied in the field of recommender systems, specifically for evaluating the performance of recommendation algorithms and mitigating or eliminating popularity bias in evaluations.




