News

Find contact details for any general enquiries.

Events

Find contact details for any general enquiries.

Student resources

Whether you're a new or continuing student, you can find everything you need here about managing your program and the opportunities available to you.

Alumni

Our alumni may be found in the world’s leading companies, policy agencies and universities.

Contact us

Find contact details for any general enquiries.

Photo: jtstockimage, Shutterstock.com

Photo: jtstockimage, Shutterstock.com

What can Twitter tell us about business?

18 December 2018

Throughout the last year, there have been many examples of business leaders and CEOs, such as Elon Musk, who have disrupted the market with no more than 280 characters. Whether through careless messages about taking a company private, or declines in a firm’s share prices after attacks from politicians, Twitter has become a very real factor in market movements.

It has also become a litmus test for public sentiment towards firms. Researchers from the ANU Research School of Accounting, Professor Susanna Ho, Dr Ka Wai Choi and Fan Yang, have published new research that explores how financial analysts are scrutinising Twitter data to detect shifts in consumer behaviour and make earnings forecasts. Using the airline industry as an example, they identify tweet characteristics and map their relationship with forecast accuracy. 

The rationale for our study is that people’s satisfaction with a firm’s service is reflected in their future purchases, which are associated with the firm’s future earnings.

“Through empirical tests, we identify associations between tweet characteristics and analyst forecast accuracy. Our findings provide support for anecdotal evidence that analysts are using opinion mining of tweets to predict firm earnings,” Professor Ho writes. 

Twitter is a trove of data and a perfect source for aggregate opinion mining. Other researchers have used it to construct linear regression models that predict the box office revenues from films, prior to their release.

This research has significant practical value to analysts, by showing the predictive efficacy of Twitter data in the domains of finance. It also highlights the value of Twitter itself as a source of data for other kinds of predictive analysis, such as politics.

“In practice, before the advent of Twitter, analysts relied on service investigation reports and expert reviews to infer consumer satisfaction about firms’ services. Analysts were unable to operate at levels of finer detail to reveal fundamental phenomena underlying the co-evolution of firms and consumer behavior.”

‘Harnessing Aspect-Based Sentiment Analysis: How are Tweets Associated with Forecast Accuracy?’ will be published in the A* Journal for the Association for Information Systems.

Read more about CBE research

The death of a close friend hits harder than we think finds new research at ANU

Find out more »

As a tool for conflict resolution, LEGO® SERIOUS PLAY® is a serious method to build shared understanding

Find out more »

There is no trustworthy evidence that manages the 'risk culture': new ANU report

Find out more »

Dr John Hewson and Danielle Wood discuss their interpretations and takeaways from the 2019 Budget

Find out more »

Updated:   22 January 2019 / Responsible Officer:  CBE Communications and Outreach / Page Contact:  College Web Team