Detecting public sentiment over PM2.5 pollution hazards through analysis of Chinese microblog

Yongzhong Sha, Jinsong Yan, Guoray Cai

Research output: Chapter in Book/Report/Conference proceedingConference contribution

7 Scopus citations

Abstract

Decision-making in crisis management can benefit from routine monitoring of the (social) media to discover the mass opinion on highly sensitive crisis events. We present an experiment that analyzes Chinese microblog data (extracted from Weibo.cn) to measure sentiment strength and its change in relation to the recent PM 2.5 air pollution events. The data were analyzed using SentiStrength algorithm together with a special sentiment words dictionary tailored and refined for Chinese language. The results of time series analysis on detected sentiment strength showed that less than one percent of the posts are strong-positive or strong negative. Weekly sentiment strength measures show symmetric changes in positive and negative strength, but overall trend moved towards more positive opinions. Special attention was given to sharp bursts of sentiment strength that coincide temporally with the occurrence of extreme social events. These findings suggest that sentiment strength analysis may generate useful alert and awareness of pending extreme social events.

Original languageEnglish (US)
Title of host publicationISCRAM 2014 Conference Proceedings - 11th International Conference on Information Systems for Crisis Response and Management
PublisherThe Pennsylvania State University
Pages722-726
Number of pages5
ISBN (Print)9780692211946
StatePublished - Jan 1 2014
Event11th International Conference on Information Systems for Crisis Response and Management, ISCRAM 2014 - University Park, PA, United States
Duration: May 1 2014May 1 2014

Publication series

NameISCRAM 2014 Conference Proceedings - 11th International Conference on Information Systems for Crisis Response and Management

Other

Other11th International Conference on Information Systems for Crisis Response and Management, ISCRAM 2014
CountryUnited States
CityUniversity Park, PA
Period5/1/145/1/14

All Science Journal Classification (ASJC) codes

  • Information Systems

Fingerprint Dive into the research topics of 'Detecting public sentiment over PM2.5 pollution hazards through analysis of Chinese microblog'. Together they form a unique fingerprint.

Cite this