Tracking labour flows with Twitter
The vast amount of data available via social networks has prompted many researchers to use that data to try and discern behavioural trends, whether it’s tracking the spread of flu or predicting stock movements.
A new paper aims to use data from Twitter to see if it can help to predict labour flows. The paper monitored Twitter over a two year period, looking out for tweets pertaining to job related situations, such as losing a job, looking for a job or a job posting.
The study made some interesting findings, and was able to do a reasonable job of correlating tweets about job losses with unemployment insurance claims, despite the obvious apparent differences between the two things. What’s more, the researchers that their social media index can also apply insights into metrics that are traditionally not so well measured, such as the number of job searches, or even job postings, both of which are very important to understanding the labour market.
The researchers are keen to point out that their solution is far from definitive, and much more work is required to make it as robust as possible, but the paper nevertheless demonstrates the kind of things that are both feasible and useful when gauging the true state of the labour market. They highlight how social data can be used either alongside or instead of more traditionally derived data, with the enormous benefit of being both low cost and in near real-time.Original post