论文部分内容阅读
在Twitter中是否存在情绪传染现象是社会科学中一个待解决的问题。首先通过LIWC2007获取了包含106 641个用户的Tw itter社交网络中所有用户的情绪时间序列,然后采用一系列的单位根检验证明了相关时间序列的平稳性,通过格兰杰因果检验,在预测用户情绪值的回归式中加入了用户关注好友过去时间的情绪值作为自变量,并采用统计假设检验的方法证明了该自变量的系数不为0,从而说明了用户的情绪会显著地被其关注好友过去的情绪所影响,即用户关注好友的情绪是用户情绪的格兰杰原因。用同样的方法证明了用户情绪并不是用户关注好友情绪的格兰杰原因,由于社交选择现象是一种双向关系,所以该结果是由情绪传染现象造成的。此外,通过统计分析发现Twitter中绝大部分单向关注好友都是非熟人,而绝大部分双向关注好友都是熟人。格兰杰因果检验的结果说明了人们的情绪既会被熟人的情绪所传染,也会被非熟人的情绪所传染。
The existence of emotional contagion in Twitter is a problem to be solved in the social sciences. Firstly, the sentiment time series of all users in Twitter social network including 106,641 users were obtained by LIWC2007, and then a series of unit root tests proved the stability of the relevant time series. By Granger causality test, Emotional value regression model adds the emotional value of the user’s past time as an independent variable, and uses the statistical hypothesis test to prove that the coefficient of the independent variable is not 0, which indicates that the user’s emotions will be significantly concerned Friends of the past, the emotional impact of the user’s attention to the mood of a friend is Granger reason the user’s emotions. Using the same method to prove that the user’s mood is not the Granger reason for the user’s concern about the emotion of the friend, the social selection phenomenon is a two-way relationship, so the result is caused by the phenomenon of emotional contagion. In addition, through statistical analysis found that most of Twitter one-way followers are not acquaintances, and most of the two-way attention to friends are acquaintances. The result of Granger’s causality test shows that people’s emotions are not only transmitted by the acquaintance’s emotions, but also by the unfamiliar emotions.