Research And System Design Of Algorithms Related To Personalized Recommendation Of College Application

The college application is an important part of the college entrance examination.However,due to the large amount of information and fluctuations of institutions and majors,it is difficult for candidates and parents to grasp the large amount of information needed for application,which leads to the phenomenon of candidates blindly applying for popular institutions and majors and others’ underwriting is very common,making candidates’ interests not match with their majors,leading to serious boredom and a surge of dropouts and academic substandard phenomena.In response to these phenomena,this paper researches the algorithm related to college application recommendation and designs and develops a college application recommendation system with personalized recommendation function based on WeChat applet platform to provide information and technical support to candidates and parents.The main research contents are as follows:(1)To address the problem that the scores of college entrance exams are influenced by many factors and the limited amount of data leads to large prediction errors,a GA-based quadratic exponential smoothing algorithm is proposed for score prediction.The quadratic exponential smoothing prediction method is chosen to predict the ranking,and the selection of the smoothing factor is optimized by using genetic algorithm.The results show that the overall absolute error of the proposed algorithm is smaller than the other two algorithms,but due to the uncertainty of the college entrance examination scores,the probability model is improved to improve the robustness of the system,and the admission information of previous years is fully utilized to reduce the dependence on the prediction results.(2)A hybrid recommendation algorithm based on user behavior characteristics is proposed for the problem of insufficient personalized recommendation ability of the college application recommendation system.For the problem that user ratings are difficult to obtain and subjective,user behaviors are mined as user ratings.Due to the problem of high dimensionality of the rating matrix caused by the large number of institutional data,the rating set is used to replace the traditional rating matrix and the Jaccard similarity coefficient is improved as the similarity calculation formula.To address the respective shortcomings of the user-based collaborative filtering recommendation algorithm and the content-based recommendation algorithm,the algorithm is mixed in a parallel manner and validated using the Ali Tianchi dataset.The results demonstrate the effectiveness of the algorithm and show the reasonableness of the representation of interest by aggregates and the improvement of the Jaccard similarity coefficient.the algorithm is applied to the recommendation system to recommend colleges and majors that suit individual interests for users with the same score and different interests,which improves the accuracy of College application recommendation.(3)In response to the problem that the diversified needs of users cannot be satisfied,an adaptive fuzzy clustering algorithm based on CLIQUE is proposed.The fuzzy clustering algorithm is used to cluster the professional data of institutions,and the priority of different classes is calculated for recommendation according to the user’s focus on different indicators.To address the problem that the fuzzy C-mean clustering algorithm cannot automatically identify the number of clusters and the initial point selection is sensitive and easy to fall into the local optimum problem,the grid clustering algorithm is used to obtain the number of clusters and the initial value,and the algorithm is verified by using test data.The algorithm computes the cluster centers with fewer iterations than the random method,which improves the computational efficiency of the algorithm.(4)In response to the fact that most of the current recommendation platforms cannot meet the timeliness and convenience of users to fill in volunteer applications,a recommendation system for college entrance examinations based on WeChat applets is developed.The information of relevant colleges and universities is organized,and the database is designed to store the data.At the same time,we analyzed the needs of candidates and parents,designed and developed the functions of the system,and implemented the recommendation algorithm in the recommendation system.The system was tested for functionality and performance,and the applet was successfully launched and promoted with about 700 users.