Opinion Matters

Team Name: 
SentiNex

This idea was born as a result of Donald Trump, Brexit and Pauline Hanson. We decided that an "out-of-touch" index what sorely needed by the voting populace to indicate how aligned government policy was with voter sentiment. We analysed refugees, detention centres and immigration - the hottest topics in the current political landscape.

The datasets we used were: the ABC Gateway API, the Australian government press releases and social media metadata related to the articles from Facebook. We scraped content data related to refugees and immigration with the use of IBM Watson technologies and applied the IBM AlchemyLab Emotion Analysis API to determine sentiment. By doing this, we then correlated the feelings of the population with the behaviour of the government (eg: "Government closes 17 detention centres"). The feelings were broken down into Joy, Anger, Fear, Disgust and Happiness, with a number assigned to each depending on the strength of each emotion. The result was an accessible, easy to understand data visualisation. Graphs, maths and a whole lot of colour.

Team Prize Details: 
University of New South Wales, Macquarie University
Used Datasets: 
Dataset Name: 
ABC Gateway API
Publishing Organisation/Agency: 
ABC
Jurisdiction of Data: 
Australian Government
Dataset Name: 
Data.gov.au
Publishing Organisation/Agency: 
Aust Govt
Jurisdiction of Data: 
Australian Government
How did you use this data in your entry?: 
The data.gov.au dataset was used to generate government press releases to run sentiment analysis on. The sentiment analysis was run with the IBM Watson EmotionAnalysis API. The sentiment of all relevant articles was then aggregated and averaged to produce a resulting sentiment output.
Event Location: 
Sydney Official