Can Twitter data be used to monitor inflation?

Jakarta has one of the largest Twitter accounts in the world – approximately 20 million user accounts. Officials in the Indonesian government realized that Indonesians tweeted a lot about food prices and that the prices mentioned in 140 characters, were often close to the official figures.

 

The United Nations Global Pulse Lab decided to make the most of this opportunity to track food price inflation in an easy and unique manner. Twitter data generated between March 2011 and April 2014 was gathered and referenced temporally, spatially and demographically. This data was then combined with public datasets concerning food and fuel prices along with various indices published by the Indonesian State Ministry of National Development Planning (BAPPENAS) and the World Food Program (WFP).

 

Using big data analysis techniques, specific words in English and Bahasa language were looked to identify tweets that were about food or fuel prices. By comparing the information derived from the many tweets and combining them with the secondary data collected, the Ministry of National Planning and Development (Bappenas) was able to come up with a model for estimating prices of various commodities. As one can see from the following graphs, the model price estimated for various commodities was very close to the official prices.

Such information can help governments track prices of various essential commodities in real time and prepare for action quickly.

 

Next time you’re out for grocery shopping, do tweet!