Distribution Characteristics Of Vocational Skills Of LinkedIn Social Network

In recent years,researches on vocational skills were very extensive.But,there was less research on the inner relationship among vocational skills based on big data.The emergence of a large number of social network and applications of social network provide the basis for our study.Such as the global social network,Linked In,has many members and holds a lot of information of vocational skills provided by members.This provides the conditions for the study of association and cluster characteristics among vocational skills.This thesis summarized the existing research results,researched the age estimation,association rules mining of vocational skills,and appraisal of vocational skills,including the following works:(1)This thesis presented a method of age prediction of members in Linked In social network based on growth model.In order to predict the age of Linked In social network members more accurately,first this method extracted the education experience and work experience of Linked In social network members,and then predicted the age of the members according to the growth model.Our method reduced the error caused by unknown image information in the age prediction method of face recognition,which greatly increased the accuracy of age prediction.(2)This thesis collected vocational skills of tens of thousands of members from Linked In social network,preprocessed and standardized these data,and then mined the association and clustering features between vocational skills.It proved that there is a large number of clustering and association relationships between professional skills,and the confidence,degree of promotion,and support of the relationship are extremely high.This shows that there is some inherent fixed relationship between vocational skills.(3)This thesis improved the distance calculation method to improve K-means++ clustering algorithm.Then,a vocational skill life curve was generated for each vocational skill,and the improved clustering algorithm was used to cluster vocational skills according to the life curves.The comparison between the generated result and unmodified clustering algorithm showed a significant improvement.The research work in this thesis can reveal the inherent relationship and dependence between different vocational skills.The research results can provide theoretical support for other aspects of vocational skills,such as vocational skills training,vocational skills mining,and appraisal of vocational skills and so on,it can also provide reference for personal employment and company recruitment.