基于行为序列分析的学习资源推荐算法研究

利用数据挖掘技术分析网络学习行为数据可以挖掘出其隐含的行为规律特征，为学习者提供个性化的学习资源服务。针对现有的数据挖掘算法在对网络学习行为数据进行分析时普遍存在模型适用性不高的问题，提出了一种基于行为序列分析的学习资源推荐算法。首先，提出行为序列及其相关概念的定义，并提出行为序列相似度计算方法；然后提出基于行为序列相似度的协同过滤推荐算法，计算学习者相似度并为待推荐学习者生成学习资源推荐列表；接着给出基于学习风格的推荐方法，将学习者学习风格特征融入推荐过程；最后，给出基于行为序列分析的学习资源推荐算法的模型。提出的算法没有对行为序列的模式进行限制，具有较高的适用性，对深入研究网络学习行为序列数据为学习者提供个性化学习服务具有一定的借鉴作用。

Research on learning resource recommendation algorithm based on behavior sequence analysis

Using the data mining technology to analyze the network learning behavior data can dig out its hidden behavior characteristics, and provide personalized learning resource services for learners. Aiming at the problem that the model applicability of the existing data mining algorithms generally was not very high when analyzing of network learning behavior data, this paper proposed a learning resource recommendation algorithm based on behavior sequence analysis. Firstly, this paper proposed the definition of behavior sequence and its related concepts. Secondly, it proposed the calculation method of behavior sequence similarity. Then, it proposed the collaborative filtering recommendation algorithm based on behavioral sequence similarity to calculate the similarity between learners and to generate the learning resource recommendation list for the learner to be recommended. And then it gave the recommendation algorithm based on learning style, and integrated learners'' learning style characteristics into the recommendation process. Finally, it presented the learning resource recommendation algorithm based on behavior sequence analysis. The proposed algorithm does not limit the pattern of behavior sequence, and has a high applicability, it can be used for further study the network learning behavior data and provide personalized learning service for learners.