Tweets are forever: a large-scale quantitative analysis of deleted tweets
|Conference paper (help)|
|Tweets are forever: a large-scale quantitative analysis of deleted tweets|
|Authors:||Hazim Almuhimedi, Shomir Wilson, Bin Liu, Norman Sadeh, Alessandro Acquisti|
|Citation:||missing booktitle : 2013|
|Publisher:||Association for Computing Machinery|
|Web:||DuckDuckGo Bing Google Yahoo! — Google PDF|
|Article:||Google Scholar PubMed|
|Restricted:||DTU Digital Library|
Tweets are forever: a large-scale quantitative analysis of deleted tweets studies deleted posts from Twitter. They try to answer what kind of differences exists between deleted and undeleted posts e.g., by using the AFINN word list for sentiment analysis.
They examined 1.6 million tweets from 292K users from a one-week period. In total they collected 67.2M tweets.
In their data set 2.4% of the tweets were deleted.
They identified two "superficial" reasons for deletions: typo or rephrasing and spam. Of the deleted tweets 17% occured because of the first reason (typo or rephrasing) and 1% was associated with spam.
They tried to examine deletion in tweets related to four selected topics: "Alcohol and illegal drug use", "Sexual activity", "Religion and politics" and "Offensive comments". They found no pattern and "roughly equal in occurence" of these topics among deleted and undeleted tweets.