社会态度反映了民众对社会的判断和感受,是社会治理需要高度重视的内容。文章对传统社会调查方法的优劣进行了分析,进而指出基于互联网计算社会态度在理论和现实层面的可行性。本文介绍了中科院心理所课题组基于网络行为的社会态度计算模型,并应用该模型计算了广东省的社会态度,通过广东省各个城市的＂地方经济满意度＂这一社会态度指标与该区域的宏观经济指标的相关性进一步验证了模型的可解释性。同时介绍了美国宾夕法尼亚大学通过Twitter数据预测公民幸福感并绘制美国的幸福感地图。利用预测模型的计算方法所采用数据真实客观,排除了社会赞许性等因素的影响,并极大地降低了成本,缩短了原有的调研周期,初步实现了社会态度的实时计算感知,对长期动态监控各项社会态度指标有极大的现实意义。希望能够与线下的社会调查法优势互补,共同为社会治理提供辅助决策。

Social governance pays much attention to requirement of all levels of society, so it is necessary to know public＇s social attitude,which reflects public＇s judgment and feeling. This paper introduced advantages and disadvantages of traditional self-report method by survey, andthen pointed out the feasibility of predicting social attitude based on Internet behavior.This paper introduces a model for predicting social attitudestudied by Institute of Psychology, Chinese Academy of Sciences. As for application, we calculated social attitude of Guangdong provinceusing the model described above. The research conducted by University of Pennsylvania predicted life satisfaction using tweets and created US wellbeing map.The data predicted has a high correlation with the result by survey. The method by using predicting model could reduce cost, ruleout the influence of social desirability, and acquire the public＇s social attitude timely, which could be a complement of conventional socialsurvey method, and mutually to be a meaningful support to public policy making.