Document Type

Thesis

Degree Name

Master of Science (MSc)

Department

Geography & Environmental Studies

Program Name/Specialization

Geomatics

Faculty/School

Faculty of Arts

First Advisor

Dr. Colin Robertson

Advisor Role

Primary Supervisor

Second Advisor

Dr. Robert Feick

Advisor Role

Member

Third Advisor

Dr. Ketan Shankardass

Advisor Role

Member

Abstract

Microblogging on geosocial platforms is a popular form of online communication where users post information about their daily lives and challenges. Since the launch of Twitter in 2006, information sharing through social media has become a largely unused data repository. Tweets often convey content about the users sentiment as it is happening. As such, Tweets can be viewed as a proxy of public mood. In this thesis, I performed a sentiment analysis of all public geo-located Tweets posted by a variety of Twitter users between September 2013 and October 2014. Each Tweet was processed through a custom algorithm to extract 8 different emotions: Anger, Confusion, Disgust, Fear, Happiness, Sadness, Shame, and Surprise. I then created an emotional landscape to display variance in emotion across the city of Toronto. The emotional landscape presented interesting emotional polarity change between the core and the periphery of the city. Neighbourhood profiles were then used to compare the emotional differences resource access could individual’s ability to cope and mediate stress. I found that individuals living within close proximity to greenspace expressed increased levels of positivity though they have decreased access to built resources. I also found that individuals within Neighbourhood Improvement Areas experienced an increased risk of negativity. I believe large-scale analyses of public sentiment can provide valuable information for further analysis of resource use in an effort to reduce negative health effects long term.

Convocation Year

2016

Convocation Season

Fall

Share

COinS