Traditional two-way human resource recommendation method has the problem of incomplete data anonymous process, which leads to low recommendation accuracy. Therefore, a two-way human resource recommendation method considering user privacy is designed. Collect recruitment information from human resources big data, and convert file formats in batches. We vectorizing each piece of information, build user privacy protection model, calculate entropy redundancy of data attribute value, optimize data anonymous process, calculate similarity between users according to preference score, and set two-way recommendation mode of human resources. Experimental results: The average recommendation accuracy of the two-way human resource recommendation method in this paper and the other two two-way human resource recommendation methods are 55.645%, 48.363% and 48.267% respectively, indicating that the two-way human resource recommendation method has a wider application scope on the basis of combining user privacy.