114.3 2009-11-11T14:41:34Z 11897 http://www.twitter.com/ 602 true Occurrence counts of tweet tokens: hashtags, URLs, & smileys by hour or month 2010-01-31T10:24:47Z bigdata,emoticon,hashtag,hours,huge,months,networking,smiley,social,socialnetwork,social_network,tokens,twitter,twitter.com,url,words 0 36 30 2009-11-11T14:56:02Z 13432 http://www.infinitemonkeywrench.com/ 2009-12-16T06:55:53Z The Monkeywrench Consultancy is an organization dedicated to producing and providing analytics for valuable data. 2009-11-11T14:56:58Z 13433 http://apiwiki.twitter.com/ 2009-12-16T06:55:57Z The Twitter API currently consists of two discrete APIs. Most application developers mix and match the APIs to produce their application. The separation of the REST and Search APIs is less than ideal and it is entirely due to history It is in our pipeline to ameliorate Twitter's API by combining the Search and REST pieces as development cycles allow. The API Overview portion of the Getting Started series explains the history. 2009-11-11T14:38:14Z 7 This data is not re-distributable in bulk form. This data may be used to produce derivative works and to power applications for commercial gain. Any API or service built on this dataset may not have the same effect as re-distributing this data in bulk. This data comes from a scrape of the Twitter social network conducted by the Monkeywrench Consultancy. The full scrape consists of 35 million users, 500 million tweets, and 1 billion relationships between users. This dataset is a corpus of tokens collected from tweets sent between March 2006 and November 2009. A "token" is either a hashtag (#data), a URL, or an emoticon (smiley face -- ;)). Think about comparing this data to the stock market, new movies, new video games, or even trendingtopics.org. For example, use it to look at the social networking adoption of Google Wave on the rate of its mentions. The tokens are binned by hour and month and the occurrence count of each token for each hour is given. The Monkeywrench Consultancy will produce custom slices, subscription services, and analysis of the full scrape for a fee. Please contact "imw@infochimps.org":mailto:imw@infochimps.org for more information. 2009-11-11T14:41:34Z 58825 2010-01-26T07:05:40Z 2009-11-11T17:02:42Z tsv 15372 41746479 602 100000 true 2009-11-20T21:12:57Z The data comes in three files: h3. tokens_by_month-20091111.tsv This file gives the occurrence count of each token for each month. It has fields * token_type -- one of "hashtag", "url", or "smiley", identifying the token type * year_and_month -- the month to which the token count refers * count -- the number of occurrences for that hour across all collected tweets * token -- the actual text of the token h3. total_tokens_by_hour-20091111.tsv This file gives the total number of tokens of each type for each hour. It has fields * token_type -- one of "hashtag", "url", or "smiley" describing the token type * date_with_hour -- the date and hour to which the count refers (in the format 2009120113, for 1pm on December 13th, 2009). All times are UTC. * count -- the number of tokens of this type for this hour h3. tweet_coverage-20091111.tsv This file gives the count of collected tweets by hour. It has fields * date_with_hour -- the date and hour to which the count refers (in the format 2009120113, for 1pm on December 13th, 2009). All times are UTC. * first_tweet_id_found -- the id of the first tweet collected during this hour. * tweet_count -- the number of tweets collected during this hour. Comparing the id of the first tweet collected in one hour and the next to the number of tweets collected in the hour gives an estimate of the coverage of the scrape during that hour. 2009-11-11T16:56:01Z tsv 15371 142531101 602 800000 true 2009-11-20T21:08:01Z The data comes in three files: h3. tokens_by_hour-20091111.tsv This file gives the occurrence count of each token for each hour. It has fields * token_type -- one of "hashtag", "url", or "smiley", identifying the token type * date_with_hour -- the date and hour to which the count refers (in the format 2009120113, for 1pm on December 13th, 2009). All times are UTC. * count -- the number of occurrences for that hour across all collected tweets * token -- the actual text of the token h3. total_tokens_by_hour-20091111.tsv This file gives the total number of tokens of each type for each hour. It has fields * token_type -- one of "hashtag", "url", or "smiley" describing the token type * date_with_hour -- the date and hour to which the count refers (in the format 2009120113, for 1pm on December 13th, 2009). All times are UTC. * count -- the number of tokens of this type for this hour h3. tweet_coverage-20091111.tsv This file gives the count of collected tweets by hour. It has fields * date_with_hour -- the date and hour to which the count refers (in the format 2009120113, for 1pm on December 13th, 2009). All times are UTC. * first_tweet_id_found -- the id of the first tweet collected during this hour. * tweet_count -- the number of tweets collected during this hour. Comparing the id of the first tweet collected in one hour and the next to the number of tweets collected in the hour gives an estimate of the coverage of the scrape during that hour. 2009-11-11T18:49:38Z tsv 15374 1479 602 true 2009-11-20T20:39:00Z Counts of all emoticons used on Twitter, March 2006 to November 2009.