Author(s):
Chao Chen* - University of South Carolina
Diansheng Guo - University of South Carolina
Abstract:
Tweets are the messages sent through twitter, the famous worldwide platform that provides an online social networking service and microblogging service. In this study, an index is proposed to describe the positivity for the state of California. The positivity index is generated from the sentiment analysis of large amount of tweets for an area. It is defined as the ratio of the positive twitter users versus the total twitter users. Positive users are people who sent more positive tweets than negative tweets in the past and negative users are defined as people who sent more negative tweets than positive tweets in the past. To analyze the spatial patterns, Kernel density maps are created to visualize the magnitude of tweets and polarity of tweets per unit area. Additionally, we applied hotspot analysis to assess the spatial clusters of the proposed index and interpreted the results with an investigation of the underlying socio-economic factors
