Replication Data for: Adaptive learning model based on collaborative filtering and deep learning

This data is from the moodle online learning platform of Peking University, and the background data is stored in the mysql database. We exported the required data from the database, stored in csv format, and encrypted the data (student number) involving personal privacy. Finally, we organized the data into three csv files, which were course information, student behavior records, and recommended video records. In addition, other data have been put together, including scoring information and other data to facilitate subsequent research. The data is described in detail below: Course Information: Record video information divided by knowledge points. Includes chapter number, chapter number, course start/end time, and course description. A total of 101 records. Promoted video: Recommended video recordings for all users. Includes user id, algorithm id, course id, and course description information. A total of 123,587 records. Student behavior information: records the user's behavior information. Where sectionid is the corresponding video number, behave is the behavior type, duration is the time from the opening of the page to the time when the behavior occurs, and the time is the standard time when the behavior occurs, and the starttime is the specific moment of the video when the behavior occurs, and the endtime is the part. The corresponding video playback time at the end of the behavior. A total of 679,079 records. The user id (student number) is processed by MD5 encryption.