2021年颁布和施行的《个人信息保护法》对个人信息处理者收集、提供、存储用户个人信息等行为进行了明确规定。在该法施行前后，各大App平台方迅速更新了自身的隐私政策。本研究通过比对《个人信息保护法》颁布前后App隐私政策完整性的变化，探究《个人信息保护法》对各大互联网平台及其App隐私政策的影响。研究分别收集了不同领域的80余款知名App在该法颁布前后的隐私政策文本，利用BERT模型搭建隐私政策文本分类模型对其进行分类预测，并结合熵权法构建隐私政策完整性评价指标体系，对App隐私政策完整性变化进行定量分析。除电子商务领域外，App隐私政策的完整性水平和分布集中程度显著提高。研究在证明《个人信息保护法》对互联网平台存在较强约束力的同时，通过构建隐私政策文本自动分类和完整性评价体系推动了隐私政策评价研究向自动化和智能化方向发展。
The Personal Information Protection Law（the PIPL）which was enacted in 2021 defines rules of collection，provision，and storage of users’ personal information clearly. After the implementation of the PIPL，most of the app platforms quickly updated their privacy policies. This study aims to explore the impact of the PIPL on integrity of Apps’ privacy policies by comparing Apps’ privacy policies before and after the promulgation of the PIPL. By collecting the privacy policy of more than 80 well-known Apps in different fields before and after the promulgation of the PIPL and using Bert model to build a privacy policy text classification model，the study successfully predicts the attributes of the applications’ privacy texts. Combined with the entropy weight method（EWM），the integrity-performance evaluation index system is constructed and used to quantitatively analyze the changes in integrity of the Apps’ privacy policy. Except for the e-commerce field，the integrity level of Apps privacy policies have been significantly improved. ［Significance］ On the one hand，this study proves that the PILP have a strong constraining force on the Internet platform. And on the other hand，the study promotes the development of privacy policy evaluation research in the direction of automation and intelligence by building an automatic text-classification and integrity evaluation system for privacy policy texts.
