量化自我App用户感知风险识别与评级方法研究  

Research on Identifying and Rating Perceived Risks of Quantified Self App Users

作　　者：李世钰 张向先[1] 闫伟 曲靖野 李贺[1] Li Shiyu;Zhang Xiangxian;Yan Wei;Qu Jingye;Li He(School of Business and Management,Jilin University,Changchun 130012;School of Information Science and Technology,Northeast Normal University,Changchun 130117;School of Computer Science and Technology,Beihua University,Jilin 132013;Library,Beihua University,Jilin 132013)

机构地区：[1]吉林大学商学与管理学院,长春130012 [2]东北师范大学信息科学与技术学院,长春130117 [3]北华大学计算机科学技术学院,吉林132013 [4]北华大学图书馆,吉林132013

出　　处：《图书情报工作》2025年第1期92-105,共14页Library and Information Service

基　　金：国家社会科学基金项目“智联网环境下用户隐私风险与隐私保护研究”(项目编号:20BTQ060)研究成果之一。

摘　　要：[目的/意义]从用户感知视角出发,提出一种量化自我App用户感知风险识别与评级方法,对于实现量化自我App用户感知风险的精确识别和有效评估具有重要的理论价值和现实意义。[方法/过程]通过构建风险内容特征词表筛选涉及风险内容的评论文本,利用Bert模型和K-Means聚类算法识别用户感知风险,并通过计算风险词频和情感倾向实现风险评级。[结果/结论]研究验证了量化自我App用户感知风险与评级方法的可行性,分析得到8类量化自我用户感知风险类别,并利用风险矩阵对其进行风险等级判断。[Purpose/Significance]This study aims to propose a quantitative method for identifying and rating perceived risks of quantitative self App users from the perspective of user perception.It has important theoretical value and practical significance for achieving accurate identification and effective evaluation of perceived risks of quantitative self App users.[Method/Process]This study constructed a risk word list to screen the review texts involving risk content,identified the perceived risks of users using the Bert model and K-Means clustering algorithm,and calculated the risk score and emotional tendency to achieve risk rating.[Result/Conclusion]The study validates the feasibility of quantifying self app user perceived risk and rating methods,analyzes 8 categories of quantified self user perceived risk,and uses a risk matrix to assess their risk levels.

关 键 词：量化自我App 用户感知风险 风险内容词表 风险识别 风险评级 

分 类 号：G203[文化科学—传播学]