Research On Learning Ability Assessment And Learning Intention Forecasting Method Based On Learning Behavior Analysis

Education informatization has entered a new stage of development and is shifting from digital education to wisdom education supported by modern information technologies such as big data analysis and artificial intelligence.Due to the characteristics of personalized service,intelligent analysis,natural interaction,and ubiquitous access,wisdom education has become the trend of the development of education informatization in China,and it is also an important measure to implement the “enhancement of education modernization” in the report of the 19 th National Congress of the Communist Party of China.At present,there are still some technical problems in the guidance,recommendation,answering,evaluation and other aspects of wisdom education.In the recommendation process,because of the semantic gap between educational resources and learners,it is difficult for users to find Learning resources needed in many educational resources.How to fill the semantic gap between educational resources and learners and achieve the semantic matching between them is the key to coping with this challenge.This paper examines two problems in how to fill the semantic gap between educational resources and learners,namely,assessing the learner's ability to learn problems and predicting learners' interest and intentional problems.The main problems that need to be solved in assessing the learner's learning ability are the accurate assessment,predicting the learner's learning ability,and the difficulty of accurately assessing the educational resources.By matching the learner's learning ability with the educational resource's difficulty level,recommend the resources of learning ability for learners,improve the learning efficiency of learners;The main problem that needs to be solved in predicting the learner's interest in learning and intent is to accurately grasp the long-term interest of the learner in certain fields and the short-term learning intentions displayed in some professions and even courses,so that the online education platform Ability to recommend to learners the educational resources they are really interested in.For the learner's problem of learning ability evaluation,this paper proposes a learning ability evaluation model based on item response theory.The model is used to obtain the learning ability value of the existing learner and tiem parameters,and then use the parameters of each item in the course to calculate the multi-level difficulty of the course,and then use these parameters to evaluate the ability of new learners.For learners with insufficient learning behavior data,this paper proposes a hybrid model combining machine learning and item response theory to solve the cold start problem of such learner's ability assessment.By combining the above three evaluation methods,we can more accurately evaluate the learner's learning ability,and combined with the multi-level difficulty of the course,we can recommend the educational resources of the corresponding difficulty to the learner through the matching of the learning ability and the multi-level difficulty of the course..For learners' interest and intention prediction problems,this paper proposes a learning interest and intention prediction model based on text analysis,and organizes some behavior data of learners in the learning platform into text data,extracting text features with improved TF-IDF algorithm and word2 vec algorithm.The extracted corresponding vectors are used to predict the learner's long-term learning interest and short-term learning intention by using the time window and introducing the attenuation factor.And for the problem that online verification cannot be carried out,another verification method is proposed to verify that the text features extracted by the model are effective,and then the learning interest and intention predicted in this paper are reasonable.