Chinese English Resume Information Extraction Method Based On HMM

In daily life,the resume has a high utilization rate.The rapid development of the Internet makes the delivery of resumes from the previous paper delivery into online delivery.With the improvement of the degree of internationalization,in the China,the use of English resume also gradually increased.Resume information extraction is an important application of text information extraction technology and also has an important application to the automatic optimization and management of resume database.Compared with the English resume written by foreigners,there are some special key fields in which written by the Chinese.The purpose of this paper is to automatically extract key fields from such resumes and through key fields to extract the information.In this paper,we establish a hidden Markov model for automatic extraction of key fields from Chinese English resumes,based on three hidden states of “Keyword”,“Non-Keyword” and “Punctuation” for word annotation.The main contents of this paper include the following item:(1)Summarize the characteristics of Chinese written English resumes.Labeling training samples manually,establishing a database of 3000 Chinese English resumes.(2)Use the established database,calculate and optimize English resumes HMM parameters.(3)This paper studies a key field extraction method based on HMM for Chinese English resumes.The model is used to automatically mark resumes with hidden state.Finally,extract the target information from the resume through the key field.The model has good performance on a database of 3000 Chinese English resumes.The correct rate and recall rate can achieve 85.99% and 83.80% in open test,as well as 85.11% and 86.03% in closed test.The experimental results verify the effectiveness and feasibility of the proposed model.