网络隐私争议事件中用户隐私关注及情感对比研究

Comparison of Privacy Concern and Sentimental Characteristics of Users in Internet Privacy Controversial Events

[目的/意义]人工智能、大数据等领城的快速发展使得商业发展与隐私保护之间的矛盾愈发尖锐。通过对不同类型的网络隐私争议事件微博评论进行情感及话题对比分析,以探究不同情境下网络用户的隐私态度的异同点与背后机理。[方法/过程]采集2012年至2019年网络隐私争议事件的相关微博评论,对其进行预处理,作为实验数据;基于情感词典计算各评论的情感强度值,并将隐私争议事件分为隐私收集类、隐私曝光类及隐私协议类,对比分析不同情境下的用户评论情感趋势;构建用户隐私讨论对象一情感表达二分网络,并通过二分网络投影构建单顶点网络,结合节点中心性等指标进行二分网络及投影分析。[结果/结论]结果表明,用户整体隐私关注呈现上升趋势;不同类型隐私争议事件的用户负面情感强度水平不同;不同隐私争议情境下用户的关注热点差异较大,情感表达各有特点。以上结果表明不同情境中的用户隐私关注及情感表现具有明显差异。

[Purpose/significance]The rapid development of AI,big data etc.makes an increasingly fierce confrontation between business development and privacy protection.This paper finds the differences in privacy con­cern and sentimental characteristics of users in different situations through the comparative analysis of sentiment and topic of Weibo comments on different types of online privacy controversial events.[Method/process]Firstly,Wei-bo comments related to privacy controversial events from 2012 to 2019 were collected and preprocessed as experimen­tal data.Secondly,the sentiment value of each comment was calculated based on the sentiment dictionary.Then,the privacy controversial events were classified into privacy collection categories,privacy exposure categories and privacy agreement categories.The sentimental analysis was conducted under different situations.Finally,a bipartite network of“discussed object-emotional expression”was constructed,and a single vertex network was constructed by the bi­partite network projection,and the analysis of both were carried out by combining the indexes such as node centrali­ty.[Result/conclusion]The result shows that users’overall privacy concern shows an upward trend;users of dif­ferent types of privacy controversial events have different levels of negative sentimental intensity;the hotspots of us­ers’concern vary widely across privacy controversial events,and emotional expression have different characteristics.These findings indicate that there are significant differences in user privacy concern and sentimental performance in different situations.