基于数据驱动和知识图谱的电影分类推荐方法

Movie Classification and Recommendation Method Based on Data Driven and Knowledge Graph

提出基于知识图谱和数据驱动的电影分类推荐方法;首先基于数据驱动爬取互联网中的电影数据并进行去重及清洗,然后采用知识图谱将电影数据与用户情感偏好数据进行关联,对海量的数据信息进行中心聚类,并在数据聚类的过程中计算投影向量得到相似度矩阵,最后查询相似度值并计算分类推荐指标权重得到最终的电影推荐清单.

A movie classification recommendation method based on knowledge graph and data-driven approach was proposed.Firstly,based on data-driven crawling of movie data from the Internet,preprocessing such as deduplication and cleaning was performed.Then,a knowledge graph was used to associate the movie data with user sentiment preference data,and a massive amount of data information was centrally clustered.During the data clustering process,the projection vector was calculated to obtain the similarity matrix.Finally,the similarity value was queried and the weight of the classification recommendation index was calculated to obtain the final movie classification recommendation list.