The Learner Model Research In The Personalized Push Service Of Education Information Resources

With the advent of web3.0, personalization and intelligentization prevail, and the individualized learning receives much attention. As a kind of personalized learning support services for the education information resources, the personalized push service can achieve a two-way matching between "resource" and "learner". And since the establition of learner model is the key procedure of the personalized push services for educational information resources, its description accuracy is directly influenceing the effect of the personalized push services. Therefore, it is of significantly theoretical and practical value to study the learner model in the personalized push service for education information resource, to analysis the learner characteristics that really influence students’ choice for resources and determine the index weight of these learner characteristics, and to research on recommendation algorithm and personalized learning services.Firstly, the thesis carries on a retrospect of the related research of learner model by literature research, through which the characteristic index that could influence learners’ preferences for educational resources in the network environment is extracted, by the help of expert interviews as well. Meanwhile the thesis use interpretative structural modeling (ISM) method to construct learners’ characteristic model, and classify the learner characteristics into five categories by hierarchy:learning progress, strategy investment, dynamic characteristics, learning style and ability. These five categories are the key elements to build learner models.Secondly, the thesis summarizes three principles that learners should follow to aquire information in network environment by the combination of explicit and implicit information, of static and dynamic information, and of general and specific information, then it probes further into specific methods to gain the different patterns of learner characteristics by means of leaner prediction and data mining. Also, the thesis introduces the analytic hierarchy process (AHP) method to establish resource selection index system on the basis of learner characteristics, and quantitatively it researches on the different learner characteristics’ degree of importance for learners’ information resource preference by Expert Investigation Method, then calculates the index weight of different learner characteristics.Finally, the undergraduate course "Selected Readings Educational Technology" is taken as an example, to research the relevance between learner characteristics and resource characteristics from side of each one of them, but separately, and to build a mapping table between learner characteristics and resource properties. Based on the personality characteristics of three learners, using Expert Investigation Method, the thesis clarifies the dominance degree of the four defferent resource alternatives for different patterns of learner characteristics, and then determines the weights of the alternatives for the different needs, thereby providing different options for the three learners.This thesis obtains information of learner characteristics by interpretative structural modeling (ISM) and builds a structure model, ie, learner characteristics model (learner learning progress, investment strategy, power characteristics, learning styles, abilities foundation); Analgtic Hierarchy Process is used to quantitatively calculate the weights of various learner characteristics, and estimate learner selection preferences for educational information resource in a relatively accurate way, in order to determine the priority of resources recommendation for improving accuracy and quality of the push service. It has certain innovation in research methods and content. However, personalized push service for educational information resources is quite a complex issue, the aquired weights of learner characteristics are, in certain sense, subjective, by the limitations in AHP method itself. The following research should explore further about the learner characteristics in network environment and the mapping relation between learner characteristics and resource properties, and obtain learner characteristics more accurately.