TY - GEN
T1 - ReadBehavior
T2 - 18th Pacific-Asia Conference on Advances in Knowledge Discovery and Data Mining, PAKDD 2014
AU - Du, Jianguang
AU - Song, Dandan
AU - Liao, Lejian
AU - Li, Xin
AU - Liu, Li
AU - Li, Guoqiang
AU - Gao, Guanguo
AU - Wu, Guiying
PY - 2014
Y1 - 2014
N2 - Along with twitter's tremendous growth, studying users' behaviors, such as retweeting behavior, have become an interesting research issue. In literature, researchers usually assumed that the twitter user could catch up with all the tweets posted by his/her friends. This is untrue most of the time. Intuitively, modeling the reading probability of each tweet is of practical importance in various applications, such as social influence analysis. In this paper, we propose a ReadBehavior model to measure the probability that a user reads a specific tweet. The model is based on the user's retweeting behaviors and the correlation between the tweets' posting time and retweeting time. To illustrate the effectiveness of our proposed model, we develop a PageRank-like algorithm to find influential users. The experimental results show that the algorithm based on ReadBehavior outperforms other related algorithms, which indicates the effectiveness of the proposed model.
AB - Along with twitter's tremendous growth, studying users' behaviors, such as retweeting behavior, have become an interesting research issue. In literature, researchers usually assumed that the twitter user could catch up with all the tweets posted by his/her friends. This is untrue most of the time. Intuitively, modeling the reading probability of each tweet is of practical importance in various applications, such as social influence analysis. In this paper, we propose a ReadBehavior model to measure the probability that a user reads a specific tweet. The model is based on the user's retweeting behaviors and the correlation between the tweets' posting time and retweeting time. To illustrate the effectiveness of our proposed model, we develop a PageRank-like algorithm to find influential users. The experimental results show that the algorithm based on ReadBehavior outperforms other related algorithms, which indicates the effectiveness of the proposed model.
UR - https://www.scopus.com/pages/publications/84901269658
U2 - 10.1007/978-3-319-06608-0_10
DO - 10.1007/978-3-319-06608-0_10
M3 - Conference contribution
AN - SCOPUS:84901269658
SN - 9783319066073
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 114
EP - 125
BT - Advances in Knowledge Discovery and Data Mining - 18th Pacific-Asia Conference, PAKDD 2014, Proceedings
PB - Springer Verlag
Y2 - 13 May 2014 through 16 May 2014
ER -