9/11/2017 11:30:00 a.m.
Prerita Gupta, a PhD student at the Complex Systems, Big Data and Informatics Initiative (CSBII) at Lincoln University and supervised by Dr Patricia Anthony and Professor Sandhya Samarasinghe, said people experiencing these upheavals turn to social media to express their feelings, especially Twitter.
She is in development of an emotion recognition algorithm to identify emotions based on tweets from 2010 to 2016 using keywords which identify those related to Christchurch’s major earthquakes and has collected more than 250,000 tweets to date.
It gauges their intense reactions after the events and how they are coping. An example is a tweet such as “Sharp lil jolt felt in this morning!!!. Forgot how much I hate them”.
“People have moved from the traditional way of expressing their feelings face to face to venting their feelings through social media as a coping mechanism, probably because they feel more comfortable sharing their feelings online,” Prerita said.
“Traditional approaches such as interviews, or surveys are not the best way to collect data as most people do not want to be reminded of the experience, as it may traumatize them further.”
She said a better way to get the information is through non-intrusive ways such as text mining approaches such as Sentiment Analysis and Emotion Identification.
Her work should have public health benefits, giving a new way to gauge emotional states.
“This information is very useful for health providers in examining the relationship between emotions and the use of anti-depressants medicines, helpful for the government to plan which affected area needs the maximum health benefit and for social welfare departments to understand the level of support they should provide to the affected party.”