基于上下文的分布式协同过滤推荐技术

针对传统协同过滤推荐技术应用于大规模动态数据集时难以兼顾准确度和效率的问题，提出一种基于上下文的分布式协同过滤推荐技术，引入推荐上下文的概念，并在此基础上充分考虑用户的即时兴趣以提高推荐的准确度，采用评分矩阵的分布式存储和计算以提高推荐的效率。实验结果表明，该分布式协同过滤技术能同时保证推荐的准确度和效率，使其在大规模动态数据集上的应用更具优势。

Context-based Distributed Collaborative Filtering Recommendation Technology

The architecture technologies of a collaborative virtual assembly system for complex product in distributed environment are studied in this paper. The system architecture supporting distributed collaborative virtual assembly is designed, and a method of supporting multi-users interactive operation in virtual assembly is proposed. Real-time rendering with high resolution for complex scene and multi-channel immersible display problems are solved. A new collaborative virtual assembly system is developed on distributed environment and multi-channel stereovision. By using the system, engineers are able to perform multi-user collaborative virtual assembly simulation for complex product easily. The validity of this system is confirmed by an example of multi-user collaborative assembly with whole vehicle.