Abstract: In the context of data being presented as a new type of production factor, more attention has been paid to the protection of personal information and the optimization of APP privacy policies. With the method of literature research, content analysis, LDA(Latent Dirichlet Allocation) topic modeling, the privacy policy framework of 200 commonly used APPs are extracted and optimized. Besides, the APP privacy policy text is analyzed from multiple perspectives and combining theoretical specifications with social practice. It is found that the current APP privacy policy framework consists of Overview of Privacy Policy, Collection of Personal Information, Use of Personal Information, Rights of Personal Information Subjects, Sharing, Transfer, Public Disclosure of Personal Information, Security of Personal Information, and Dispute Resolution of Personal Information Protection, which can be optimized from a two-dimensional structure to a three-dimensional frame.

Key words: Privacy policy, Personal information, Latent Dirichlet Allocation, Text analysis, Mobile application；Privacy protection；APP