| 摘 要: 为了提高大数据中动态用户个性化推荐的准确性和效率，采用基于混合协同过滤的方法来完成用户感 兴趣数据的筛选，从而实现个性化推荐。先将用户数据及项目数据通过协同过滤算法来完成建模并评分，然后结合 XGBoost模型的树形结构和正则学习的特点进行预测评分，接着将两种算法混合来求解最优目标函数，得到候选的推荐 数据集合。最后通过实例仿真，混合算法精确度高，在大数据平台有较强的适用性。 |
| 关键词: 大数据；协同过滤；XGBoost；个性化推荐；准确率 |
|
中图分类号: TP399
文献标识码: A
|
|
| Personalized Recommendation for Dynamic Users Based on Hybrid Collaborative Filtering in Big Data |
|
LIU Shanshan
|
( South China Vocational College of Commerce and Trade, Guangzhou 510650, China)
|
| Abstract: In order to improve the accuracy and efficiency of personalized recommendation for dynamic users in big data,a hybrid collaborative filtering method is used to filter the data of interest to achieve personalized recommendation. Firstly,user data and project data are modeled and scored by collaborative filtering algorithm,then predicted and scored by combining the tree structure of XGBoost model and the characteristics of regular learning.Then the two algorithms are mixed to solve the optimal objective function,and candidate recommended data set is obtained.Experiments show that the hybrid collaborative filtering recommendation algorithm has high accuracy and strong applicability in big data platforms. |
| Keywords: big data;collaborative filtering;XGBoost;personalized recommendation;accuracy |
