I am a newbie. I would like to do visualization on twitter data : top trends based on country (over map) and time variations (for each months in 1 year or for each year) . Can someone tell me where can I get the twitter data set and any advice on how to start proceeding would be really help full.
Thanks
(Fresh and complete) Twitter data is not easy to get nowadays, as Twitter is actively monetizing it (through GNIP). There is still a large Twitter graph collected by Kwak et al. for their "What is Twitter, a social network or a news media?" WWW'10 publicly available, but as to tweets, most nice datasets have disappeared from the Internet due to complaints from Twitter. There are still tiny datasets of tweets available online (e.g.), but they are small and stale. Thus, I see two "legal" ways to go about getting tweets now: 1) crawl Twitter using the new version of Twitter API (the obvious limitation of this approach is the limit on the number of requests you can send; this approach is unlikely to give you a representative picture of what is trending if you collect raw tweets, just because you won't have enough tweets); 2) buy data from Twitter (GNIP), which will cost an awful lot of money. Alternatively, you can 3) study Twitter API for other means of getting "summarized" information from it, not tweets directly. For example, they may provide a method to retrieve, say, 10 most popular hash tags at the moment, maybe, even filtered by geo-location -- you cannot do much data science with the result, but at least you can use it in your visualization application.
There's a tweet collection (comprising of 16 million tweets) that you can get from NIST. This collection is used for the ad hoc search task on micro-blogs. TREC 2011 Microblog collection
