twitter

**** CITATION ****

Please cite our paper as follows, when you are using our dataset:

Lilian Weng, Filippo Menczer, and Yong-Yeol Ahn. Virality Prediction and Community Structure in Social Networks. Nature Scientific Report. (3)2522, 2013.

**** DATA SOURCE ****

Sampled public tweets from Twitter streaming API (https://dev.twitter.com/docs/streaming-apis).

Date range: March 24, 2012 to April 25, 2012.

**** NETWORKS ****

follower_gcc.anony.dat:

    Format: anony.user1.ID anony.user2.ID

    Anonymized reciprocal follower network.

    Each edge is a pair of Twitter user who are following each other. After recovering the reciprocal follower network, the giant connected component is extracted.

retweet_gcc.anony.dat:

    Format: anony.user1.ID anony.user2.ID weight

    Anonymized reciprocal retweet network.

    Similarly to follower_gcc.anony.dat, but instead each edge is a pair of users who retweeted each other at least once during our observation time window. Weight is the sum of how many times user1 retweeted user2 or user2 retweeted user1.

mention_gcc.anony.dat:

    Format: anony.user1.ID anony.user2.ID weight

    Anonymized reciprocal retweet network.

    Similarly to follower_gcc.anony.dat, but instead each edge is a pair of users who mentioned each other at least once during our observation time window. Weight is the sum of how many times user1 mentioned user2 or user2 mentioned user1.

**** HASHTAG SEQUENCES ****

timeline_tag.anony.dat

    Format: hashtag timestamp1,anony.user1.id timestamp2,anony.user2.id ...

    Each line is a hashtag followed by the sequence of its adopters sorted by timestamp. A user is considered as an adopter of a hashtag once he/she starts using the hashtag. We only consider users who appear in the collected networks. The timestamp is the time when we see the hashtag in the user's tweets. The file includes both emergent hashtags and non-emergent ones.

timeline_tag_rt.anony.dat

    Format: hashtag timestamp1,anony.retweet_user1.id,anony.retweet_from_user1.id timestamp2,anony.retweet_user2.id,anony.tweet_from_user2.id ...

    Each line is a hashtag followed by the sequence of its adopters retweeting about this hashtag from other users sorted by timestamp. A "retweet_user" retweets a message containing the hashtag from a "retweet_from_user". We only consider users who appear in the collected networks. The file includes both emergent hashtags and non-emergent ones.

timeline_tag_men.anony.dat

    Format: hashtag timestamp1,anony.mention_user1.id,anony.mentioned_user1.id timestamp2,anony.mention_user2.id,anony.mentioned_user2.id ...

    Each line is a hashtag followed by the sequence of its adopters mentioning other users in messages containing this hashtags sorted by timestamp. A "mention_user" mentions a "mentioned_user" in a message with the target hashtag. We only consider users who appear in the collected networks. The file includes both emergent hashtags and non-emergent ones.

** Note that users in these networks and timeline sequencies are anonymized in the same way so that the same IDs refer to the same Twitter users.