Assessment Of Online Learning Effectiveness By Integrating Learner Implicit Feedback

At present,online learning effectiveness evaluation is mainly based on explicit data such as test scores and review texts.There are some problems in the evaluation model,such as one-sided evaluation basis,ignoring the relationship between learning behavior and so on.In order to solve these problems,this thesis proposes and implements a learning effect evaluation model based on frequent itemsets of learning behavior,which integrates implicit feedback data.The main work of this thesis is as follows:Firstly,the characteristics of implicit feedback are introduced.Twelve kinds of implicit feedback data sources in online learning enviroument are proposed,and the method of collecting implicit feedback data through embedded JS script is given,which improves the diversity of evaluation data sources.Secondly,aiming at the problem that the frequent itemset mining algorithm will produce tens of thousands of frequent patterns when the support degree is low,an improved grouping strategy is proposed,which realizes the parallelization of the algorithm,and the implementation method on Spark platform is given,which provides an effective support for mining the relationship between learning behaviors.Then,according to the characteristics of implicit feedback data in online learning scenarios,the specific methods of data acquisition are given and appropriate preprocessing is carried out.The LBPFP algorithm proposed in Chapter 3 is used to mine frequent itemsets and find out the relationship between learning behaviors.Based on this,a scientific and reasonable learning effect model is constructed.Finally,this thesis realizes the evaluation system which integrates the implicit feedback data of online learners,gives the system structure diagram,introduces three main modules:data acquisition and preprocessing module,model training module and effect evaluation module,and shows the good operation effect of the system.