How to get twitter historical data?
I am doing a research in twitter sentiment analysis related to financial predictions and i need to have a historical dataset from twitter backed to three years. last year twitter announced that they will release historical data for scientific proposes.
I am asking if anybody have an idea about how to get this data?
All Answers (6)
Last thing I know, Twitter released the full historical data per user, but I'm really interested in this question. Good luck!
I am also very interested in knowing the answer to this question. So good luck in hunting down information!
As far as I know, there is no way to do it for free. These guys provide the data: https://gnip.com/, the prices start from 1000$. You can write them and ask how much it would cost for the data you need.
I've used Amazon AWS and among various plug-in there was also a twitter plugin to download posts and analyze them. Take a look at: http://docs.aws.amazon.com/gettingstarted/latest/emr/getting-started-emr-sentiment-tutorial.html. The price depend on the time that you consume in AWS cloud. Anyway it wasn't so much expansive. With a voucher of 100$ you can play a lot with it.
You can can do it most inexpensively for smaller jobs via Sifter, which is a self-serve web service with unlimited free estimates and powerful filter rules, provided by a small business member of the Plugged In to Gnip partner program. If you make a purchase, the data is placed in an advance text analytics platform (DiscoverText.com), where you can analyze it for free using search, filters, duplicate detection, near-duplicate cluster, human coding, and machine learning.
Only way to get access to historical data is you have go through Gnip. As Lidia mentioned check the cost for what you need.
I too had a similar problem and because of the cost had to rely on current tweets instead of historical ones and downloaded them manually.
https://gnip.com/products/historical/