移动应用隐私权声明内容合规性检验方法

The privacy policy statement of a mobile application serves as a crucial document that must be disclosed to users before collecting their information. However, current privacy policy statements face various issues, such as missing key disclosure items, omitting information collection purposes, and using vague descriptions. With an increasing number of legal provisions, the requirements for privacy policy statements vary, making compliance verification more burdensome. This study proposes a multi-label classification method for mobile application privacy policy statements. This method compares the requirements of four core laws and regulations regarding privacy policy statements, summarizes and organizes 31 categories of core item labels and features. Under this label system, the study designs and implements a classification model for privacy policy statement sentences, which achieves a 94% accuracy rate in item classification. Using this model, compliance verification was conducted in Android applications and mini-program scenarios, revealing issues such as missing items (79%), omitted purposes (63%), and vague descriptions (94%) in privacy policy statements.