The social media data’s usage has been explored in various ways over the years. The usage ranges from health issues to political campaigns. Now a study claims that social media data can be used to help predict crimes.
The study was conducted by four researchers from the University of California, Los Angeles, and the University of Victoria, Canada. Sean Young, the executive director of the University of California Institute for Prediction Technology, led the team. His other three team members were Yan Wang and Wenchao Yu of UCLA and Sam Liu of the University of Victoria.
From May 26 to Dec. 9, 2012, the researchers collected 553,196,061 tweets, using Twitter’s free advanced programming interface (API) and the “garden-hose” method. The tweets contained keywords related to crimes and having geo-locations. Later the tweets were analyzed.
The study found that social media can be a useful tool in monitoring crime. “Our findings provide evidence that it is possible to use social media in addition to existing tools to monitor crime,” the study said. It also suggests that in the future, such activities to monitor and predict crimes would be more accurate.
The study also revealed a correlation between county-level crimes and drug-related tweets.
“We found that drug-related tweets in 2012 were associated with county-level crime rates in 2012 and 2013,” the study concluded.
The study, however, mentioned that the way of monitoring and predict crime through Twitter would have benefits as well as drawbacks. The drawbacks are “ethical implications” of such surveillance and high expense to implement the program, while the high availability of the data would be considered as the benefit of the study.
To read more: https://doi.org/10.1177/2056305119834585
Wang, Y., Yu, W., Liu, S., & Young, S. D. (2019). The Relationship Between Social Media Data and Crime Rates in the United States. Social Media + Society. https://doi.org/10.1177/2056305119834585