摘要
在传统推荐系统模型的基础上,结合预购买家纺产品的特性,建立了预购买家纺产品的推荐模型。研究了推荐系统预购买家纺产品的偏好度的数据提取算法。在此基础上,结合用户在购买家纺产品过程中具有确定性与不确定性的特性以及模糊神经网络的特性,提出了一种模糊神经网络推荐算法。通过对推荐系统的算法的量化研究、仿真研究,显示算法具有较小的MAE值,且随着预购买家纺产品采样人数的增加,MAE依然保持较小的数值,证明了推荐系统的基准集与推荐集具有较好的一致性、稳定性,推荐系统具有较好的聚类特性,说明了基于模糊神经网络算法在家纺产品电子商务推荐系统是可行的。
Recommendation model to buy textile products used by family is established after the traditional recommend system model is analyzed,combining to buy the textile products characteristics.First,it analyzes data collection and data refining algorithm of recommender system to buy textile products used by family.Then a fuzzy neural network recommendation algorithm is built,which is based on fuzzy neural network features and on fuzzy to buy textile products.Finally,simulation experiment show that these values of MAE is quite small,and keeping these small values of MAE with sampling data increasing.All these results demonstrate that the recommendation algorithm performed well.
出处
《计算机技术与发展》
2011年第9期183-186,共4页
Computer Technology and Development
基金
浙江省教育厅科研项目(Y201017814)
宁波市自然科学基金项目(201001A6001043)
关键词
推荐系统
模糊神经网络
聚类
MAE
仿真
recommendation system
fuzzy neural network（FNN）
clustering
MAE
simulation
