Chinese Resume Information Extraction And Recommendation

Online recruitment has become the most popular recruiting methods due to its low cost,convenience and other advantages,but it could not automatically parse and recommend resumes.In order to solve the problem of low efficiency of selecting resumes manually.This thesis puts forward a new solution of automatic resume parsing and recommender algorithm of Chinese resumes to select suitable candidates for recruiters.The contents of this thesis include:(1)According to the hierarchical structure of the Chinese resume,the segmentation of Chinese resume is transformed into the classification problem,and all the paragraphs in Chinese resumes are classified into six predefined general classes,such as personal basic information,job hunting intension,working experiences and so on.(2)This thesis proposes an information extraction method of Chinese resume based on statistical models and rules,which considers the strong regularity and statistical characteristics of Chinese resumes.It can not only extract the simple information like name,gender,and contact information from personal basic information blocks and other simple information blocks,but also extract complex information using HMM model,such as work experience,which is attractive and useful for recruiters.(3)This thesis proposes a content-based reciprocity recommender(CBRR)algorithm to achieve the automatic recommendation of resumes.The experimental results show that the CBRR algorithm is better than the common collaborative filtering algorithms.At the same time,the CBRR algorithm is better than other reciprocity recommender algorithms because the CBRR algorithm considers the matching scores of skills and job responsibilities.(4)Designing and implementing a Chinese resume parsing and recommender system which is based on the B/S architecture,in which the rules and statistics based information extraction algorithm and CBRR algorithm are applied to the practice.