I am currently undertaking a study which uses an innovative cloud-computing tool – developed here at SMART Infrastructure Facility – to collect, analyse, and visualise data from different social media networks in Australia, China and Korea. Specifically, it aims to investigate how different countries are using social media platforms respectively to contribute to topical conversations on culture and politics. By gathering, sorting, and displaying information from millions of data records on various topics such as films and TV dramas, this project will increase public awareness among three different countries.
To increase deeper mutual understanding of the cultural and political issues that matter to each country, a range of text records (and their corresponding geography data) will be collected and placed into databases that can be used for analysis in a number of different ways:
- Supervised and Unsupervised machine learning techniques – to assist with classification, clustering and prediction;
- Geospatial analysis to search for clusters of texts on particular topics in geographical regions;
- Time series analysis to examine trends of these messages over particular periods;
- Sentiment analysis to discover the attitudes of authors of these social media texts;
- Graph analysis to understand social connections between authors (producers) and consumers (readers) of these social media texts.
In summary, this project will facilitate the application of a useful big data technique for exploring how social media are shaping public opinion.