TY - GEN
T1 - A text cube approach to human, social and cultural behavior in the Twitter stream
AU - Liu, Xiong
AU - Tang, Kaizhi
AU - Hancock, Jeffrey
AU - Han, Jiawei
AU - Song, Mitchell
AU - Xu, Roger
AU - Pokorny, Bob
PY - 2013
Y1 - 2013
N2 - Twitter is a microblogging website that has been useful as a source for human social behavioral analysis, such as political sentiment analysis, user influence, and spread of news. In this paper, we discuss a text cube approach to studying different kinds of human, social and cultural behavior (HSCB) embedded in the Twitter stream. Text cube is a new way to organize data (e.g., Twitter text) in multiple dimensions and multiple hierarchies for efficient information query and visualization. With the HSCB measures defined in a cube, users are able to view statistical reports and perform online analytical processing. Along with viewing and analyzing Twitter text using cubes and charts, we have also added the capability to display the contents of the cube on a heat map. The degree of opacity is directly proportional to the value of the behavioral, social or cultural measure. This kind of map allows the analyst to focus attention on hotspots of concern in a region of interest. In addition, the text cube architecture supports the development of data mining models using the data taken from cubes. We provide several case studies to illustrate the text cube approach, including public sentiment in a U.S. city and political sentiment in the Arab Spring.
AB - Twitter is a microblogging website that has been useful as a source for human social behavioral analysis, such as political sentiment analysis, user influence, and spread of news. In this paper, we discuss a text cube approach to studying different kinds of human, social and cultural behavior (HSCB) embedded in the Twitter stream. Text cube is a new way to organize data (e.g., Twitter text) in multiple dimensions and multiple hierarchies for efficient information query and visualization. With the HSCB measures defined in a cube, users are able to view statistical reports and perform online analytical processing. Along with viewing and analyzing Twitter text using cubes and charts, we have also added the capability to display the contents of the cube on a heat map. The degree of opacity is directly proportional to the value of the behavioral, social or cultural measure. This kind of map allows the analyst to focus attention on hotspots of concern in a region of interest. In addition, the text cube architecture supports the development of data mining models using the data taken from cubes. We provide several case studies to illustrate the text cube approach, including public sentiment in a U.S. city and political sentiment in the Arab Spring.
UR - http://www.scopus.com/inward/record.url?scp=84874814009&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84874814009&partnerID=8YFLogxK
U2 - 10.1007/978-3-642-37210-0_35
DO - 10.1007/978-3-642-37210-0_35
M3 - Conference contribution
AN - SCOPUS:84874814009
SN - 9783642372094
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 321
EP - 330
BT - Social Computing, Behavioral-Cultural Modeling and Prediction - 6th International Conference, SBP 2013, Proceedings
T2 - 6th International Conference on Social Computing, Behavioral-Cultural Modeling and Prediction, SBP 2013
Y2 - 2 April 2013 through 5 April 2013
ER -