Research On Learning Intention Recognition Based On Learner Behavior Analysis And Educational Resource Recommendation

In recent years,online education and MOOC have emerged,and the "Internet +Education" model has been widely accepted and approved.Various education platforms have emerged one after another.With the continuous improvement of the online education and teaching system,more and more emerging courses have poured into online education platforms,the learner groups targeted by online education have also continued to expand.Online education resources have become huge in number and diverse in types,making the efficiency of learners’ retrieval of resources inefficient,resulting in insufficient use of high-quality educational resources.Learning intention is a potential factor that determines how learners choose courses,how to learn,and other learning behaviors.Therefore,learning intention identification is of great value to problems such as perfecting learner profile and optimizing educational resource recommendation.Based on above research background,this paper analyzes the learning behavior data of learners in online education platforms,and defines and classifies learning intentions.Since educational resources include three levels,knowledge,courses,and domains,and learners’ learning behaviors have their own characteristics in each of these three levels,this paper divides learning intentions into knowledge learning intentions,course learning intentions,and domain learning intentions,using KMeans clustering,LSTM+Attention model and index analysis method to analyze.At the same time,this paper established a course learning intention data set by manual labeling,which contains learners’ learning records in their selected courses and corresponding intention labels.In terms of education resource recommendation,this paper analyzes the advantages and disadvantages of the existing recommendation technology,as well as the degree of matching with the research content.And then according to the characteristics of online education resources,this paper proposes a video recommendation combining collaborative filtering and knowledge learning intention clustering method,and a course recommendation method based on the field-aware decomposition machine’s integration with domain learning intention.In the research process of this subject,the real data set provided by the research group was used to verify the experimental effect after preprocessing.Experiments show that the definition,classification,and recognition methods of learning intention proposed in this paper are reasonable.And the recommendation model that incorporates learning intention is slightly improved compared to the traditional model.Finally,a learning intention visualization and educational resource recommendation system is designed and implemented by applying the research method.