中文摘要：

个性化推荐是目前解决电子商务中产品信息过载问题的有效工具之一。对综合用户偏好模型和BP神经网络的个性化推荐算法进行了研究。具体讨论了如何建立用户偏好模型，采用神经网络训练得到目标用户的偏好模型，通过Movielens数据库验证该模型的有效性。提出了一个基于内容的个性化推荐算法。

英文摘要：

Personal recommendation is very effective to find the useful information from database of products for customers in electronic commerce. The paper investigates personal recommendation algorithms based on customer preference model and BP neu- ral networks. In details, a customer preference model is proposed and BP neural network is used to train the model. Movielens data- base is used to verify the validity of BP neural network model. A content-based personal recommendation algorithm is proposed.