University Employment Recommendation Method Based On Implicit Association Rules Ontology Model

University graduates are the most important group in the labor market,and their employment situation has an important impact on economic development and family happiness.At present,the idea of collaborative filtering is widely used in the field of employment recommendation.Because of cold start and sparse scoring records,it will lead to unmatched recommendation,which will restrict the further improvement of recommendation accuracy.There are blindness,individual differences and cognitive limitations in the employment decision-making process of university graduates,but there are also certain clusters.At present,although the research on employment recommendation for university graduates has made many innovations in the algorithm,it has not fully utilized the influence rules of graduates’ personal attributes on employment decision-making.It has mainly carried out research from the similarity calculation of user rating prediction and resume description information,which has strong subjective dependence and poor reliability.The employment recommendation of university graduates was taken as the research object in this thesis takes,aiming at the problems of cold start and low recommendation accuracy in employment recommendation,and the recommendation methods were presented for two main employment destinations: employment and further studies.This method adopted content-based recommendation algorithm to alleviate the cold start problem.On this basis,an implicit association rule ontology model was proposed and used for recommendation to improve the recommendation accuracy.The main work was as follows:(1)Employment recommendation method based on implicit association rules ontology model.Firstly,the FP-Growth algorithm was used to mine the association rules of graduates’ attributes and employment company attributes,and then the ontology model was constructed to analyze graduates’ preference characteristics.Secondly,the feature concept was defined,the preference vector was constructed,the ontology model was embedded into the content-based recommendation algorithm,the similarity between the preference vector and the company attribute was calculated by the multi-source preference ontology compatibility module,and the weighted similarity was calculated according to the association strength in the ontology model;The hidden regional flow bias module was adopted to search the neighborhood job;The recommended sequences are connected in series and bilateral conditional matching was carried out through the bilateral matching module.Finally,the importance of each module in this method and the feasibility and effectiveness of this method were verified by ablation experiments and comparative experiments of different disciplines and methods.(2)Recommending method based on implicit association rules ontology model.Firstly,the feature concept was defined according to the selection preference feature of universities for further studies,the implicit association rule ontology model in the field of university recommendation was built,and the multi-source preference ontology compatibility module was designed to obtain the recommendation sequence of graduates’ feature preference;Secondly,in order to adapt to the preferences of students with different learning abilities and backgrounds,a multi-factor weight attribute matching module was designed,and the weighted similarity was calculated according to the importance of different factors,and the groups with high weighted similarity were screened,and the similar groups’ colleges and universities were connected in series with the feature preference recommendation sequence to form a recommendation sequence;Finally,the validity of the method was verified in the real survey data.The method proposed in this thesis could provide a new way of thinking for the study of employment recommendation,and a methodological basis for the design of employment recommendation system in universities.