基于云计算的知识服务推荐系统研究

发布时间：2018-11-28 07:34

【摘要】：推荐系统可以根据用户信息及行为，例如性别、年龄、爱好以及用户选择记录，从海量知识中选择其可能感兴趣的内容推荐给用户。推荐系统很好的满足了知识服务的个性化服务特征。推荐系统对用户信息和行为数据的不断采集，推荐质量也在相应提高，不断接近精确推荐。推荐系统在社交网络、电子商务、搜索引擎、互联网广告营销中具有重要的意义。今天，我们的学习系统与社交网络、搜索引擎等密不可分，那么研究推荐系统对促进我们的学习也具有重要的意义。云计算平台为推荐系统提供了天然优势。首先，云中的数据存储是集群化的，存储管理是虚拟化的，理论上为推荐系统提供了无容量限制的数据存储能力和高效的数据吞吐能力，推荐系统因此可以拥有能快速获取、海量的训练数据，得以提供优质的推荐结果；其次，，云的分布式计算能力和物理资源虚拟化为推荐系统提供了较高的响应能力，这有助于为大量用户提供个性化推荐。通过对知识服务、推荐系统、云计算相关技术的阐述，构建了个性化推荐系统模型，构建了云环境下的知识库，构建了用户模型，并在MapReduce的基础上改进了基于协同过滤的推荐算法，使推荐系统适应当前海量数据时代的计算要求。在理论上对云计算环境下的知识服务具有一定的探索意义，在实践上对向学习者推荐个性化知识服务具有参考价值。
[Abstract]:According to the user information and behavior, such as gender, age, hobbies and user selection records, the recommendation system can select the content of its possible interest from a large amount of knowledge to the user. The recommendation system satisfies the personalized service characteristics of knowledge service. With the continuous collection of user information and behavior data, the recommendation quality is also improved, and the recommendation system is close to accurate recommendation. Recommendation system is of great significance in social network, e-commerce, search engine and internet advertising marketing. Today, our learning system is closely related to social networks and search engines, so it is of great significance to study the recommendation system to promote our learning. Cloud computing platform provides the natural advantage for recommendation system. First of all, data storage in the cloud is clustered and storage management is virtualized, which theoretically provides the recommendation system with unlimited data storage capacity and efficient data throughput, so the recommendation system can be quickly acquired. Massive training data to provide quality recommendations; Secondly, the distributed computing power of cloud and virtualization of physical resources provide high response ability for recommendation system, which is helpful to provide personalized recommendation for a large number of users. Through the elaboration of knowledge service, recommendation system, cloud computing related technology, the personalized recommendation system model is constructed, the knowledge base under cloud environment is constructed, and the user model is constructed. The recommendation algorithm based on collaborative filtering is improved on the basis of MapReduce, so that the recommendation system can meet the requirements of computing in the age of mass data. In theory, it has certain exploration significance to knowledge service in cloud computing environment, and has reference value to recommend personalized knowledge service to learners in practice.
【学位授予单位】：河南师范大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：TP391.3

【参考文献】