Research On Evolution Modeling And Analysis Of Academic Ability Of College Students

Higher education aims at the training of high-level talents,teaching,research and social services.Its significant mission lies on cultivating senior professionals,developing science and technology,inheriting excellent culture and promoting socialist modernization.Of these issues,the kernel essence comes to the enhancement of talent competency in the process of the development and reform of higher education.With the development of information technology,the boom of novel approaches and means of pedagogy comes to the field of higher education.Great changes of learning styles occur to the college students.However,the student's state and progress of learning change from individual to individual.They will be affected due to individual attributes and attitudes as well as the comprehensive influence of academic environment and others.At present,substantial limitations are shown in the study of evaluating student's learning status and process.Typically,they are focusing on the learning performance of students themselves.There are few studies on the environment and the evolutionary process which students involve.They exhibit less instructive effect on the students.Nonetheless,the existing evaluation methods are still inadequate.Based on the learning data of higher education students,this thesis studies the evolutionary model of student's academic ability.It focuses on the evolution method of academic ability,constructs the academic ability evolution model based on heterogeneous information network,and excavates the evolution law of academic ability group and individual evolution law.At the same time,a ranking method based on the machine learning is adopted to model student's academic performance dataset to explore the connection between academic ability elements and academic ability results.Based on the learning ranking learning model,the learning situation of college students is predicted,and the teaching effect is reasonably evaluated and continuously improved according to the feedback of graduates.Based on this basis,the academic ability analysis framework of "ability evaluation and feedback improvement" is proposed,and the corresponding system prototype is constructed.The main research contents are as follows:(1)The effective sampling of data from dynamically complex academic competence evolution processesThis thesis proposes an adaptive data selection mechanism of learning ability under the considerations of issues like multiple dimensions and large data characteristics of student's learning process as well as the considerations of the relevant influence and association of individual middle school students in the learning process.The proposed mechanism is built on the model of heterogeneous information networks and effectively samples the evolutionary process data using adaptive data sampling methods as the basis for evolutionary modeling.Experiments show that the proposed method can effectively select representative data during evolution and provide effective support for academic ability modeling.(2)The modeling of group evolution of academic abilitiesThis thesis investigates the factors of influencing the evolution process in the complicated environment of the dynamic academic ability.It puts forward a mining method on academic ability group evolution.This method analyzes the association of communities in different evolution stages,connects the community before and after different stages,and mines the change law of community in evolution.Experiments show that the proposed evolutionary modeling method can effectively represent the evolution of the student population.(3)The analysis of student academic ability based on ranking of learningThis thesis examines the evaluation issue of undergraduate student's academic scores in the colleges.The professional achievements are used as the feedback.Machine learning technologies are introduced into academic performance ranking task.The analysis framework of learning ability is created through combining the feedback learning ranking model.At the same time,this thesis explores the influence of postgraduate professions among different student groups about the academic commonness and differences.A series of comparative results on the real dataset verify that the proposed model has good prediction accuracy and can achieve reasonable ranking by academic performance and thus be used to improve instructional evaluation activity.(4)The academic ability analysis and modeling using "ability evaluation and feedback improvement"This thesis builds the framework of academic ability evolution by modeling and analysis based on educational data.The framework "ability evaluation and feedback improvement" includes two parts.The first one is to construct the model of academic ability evolution;the other is to complete the academic ability analysis based on the feedback leading sequencing learning.The former is the positive ability evaluation part,including the division of student's academic ability network evolution stage and data selection and the evolution of academic ability network.The latter is the feedback improvement part,which mainly analyzes the influence of academic ability factors through the ranking model to predict the academic ability in study.Based on the full discuss on the above research contents,this thesis analyzes the influence of various environmental factors on the student's academic ability group and individual evolution law,constructs the academic ability evolution model and analysis model and applies it to the actual teaching scene.The experimental results show that the proposed algorithm can achieve relatively good realization accuracy and provide effective support for student's learning process evaluation and college undergraduate teaching quality evaluation.