Research And Implementation On Key Technologies Of The College Entrance Intelligent Application System

Higher education enrollment aspiration filling is an important part of the college admissions.However,the massive amount of college entrance examination information and short application time make it difficult and complicated.Higher education enrollment aspiration filling is in urgent need of information management and intelligent transformation.Comprehensively processing the relevant data of the college entrance filling and realizing the personalized recommendation and review of the candidate's college entrance filling form is an important part of the informatization and intelligence of the college entrance enrollment aspiration filling.Comprehensively processing the relevant data of the college entrance filling and realizing the personalized recommendation and review of the candidate's college entrance filling form is an important part of the informatization and intelligence of the college entrance enrollment aspiration filling.Therefore,it is of great significance to design and develop a set of intelligent filling system for college entrance application.This project takes the construction of the college entrance Intelligent Application model as the research goal,and uses machine learning and personalized recommendation technology as the means to carry out research work around the college admission line prediction and college entrance filling recommendation.The main research contents of the thesis are as follows:(1)A combination prediction model of college admission scores based on Support Vector Regression(SVR)is constructed.Aiming at different factors that affect college admission scores,the linear regression prediction method and the difficulty coefficient prediction method are used to construct the first-level prediction model.The result of the first-level model is used as the input feature of SVR,and a combination prediction model of college admission scores based on SVR is constructed.Experimental results show that this model performs well on different experimental data sets.The mean absolute error of prediction is 6.41,5.26,11.47,which are all lower than the single prediction model and other machine learning models,which proves that this method can effectively reduce the prediction error and improve Forecast accuracy.(2)A college admission score prediction model based on Two Way Multilayer Percetron(TWMLP)is constructed.Considering the factors such as the college's development level,the college's admission scores,etc.Divide the characteristics of colleges into two passes through the multi-layer perceptron model to predict college admission scores respectively,and construct the final model by dynamic weighting method.The model has achieved good performance in the selected data set.The mean absolute error is 5.72,the root mean square error is 10.39,and the goodness of fit is 0.963.Compared with other models,this model maintains the high prediction accuracy while reducing the overall prediction error of the model,improves the goodness of model fit,solves the problem of large single feature prediction error and unstable prediction effect,and enhances the robustness of the model.(3)A college entrance filling recommendation model incorporating intelligent review is constructed.Based on the strategy of college entrance application and the idea of gradient division,intelligent analysis and reasonable evaluation of candidates' college entrance filling are carried out,unreasonable college entrance filling forms are screened,and the problem is pointed out.Combined with the candidate's initial college entrance filling form,the candidate's personalized label is extracted,and personalized college entrance application recommendation is implemented according to the candidate's preference.It is proved by examples that this model provides new ideas for solving the problem of rationalization of candidates' college entrance application,and effectively reduces the risk of the college entrance application.(4)The design and implementation of the intelligent filling system for college entrance Application.Combining the research results of this subject,using PHP language,using MVC design pattern and thinkphp 5.0 framework to design and develop a college entrance intelligent application system that integrates information comprehensive retrieval,the personalized recommendation of the college entrance application and the intelligent review.of the college entrance application.To help candidates quickly grasp a large amount of information,scientifically plan the college entrance Application.At present,the college entrance intelligent application system has been registered and used by more than 8,000 users,and the services provided have received unanimous praise.