TY - GEN
T1 - An incremental anytime algorithm for mining T-patterns from event streams
AU - Johnson, Keith
AU - Liu, Wei
PY - 2018/1/1
Y1 - 2018/1/1
N2 - Temporal patterns that capture frequent time differences occurring between items in a sequence are gaining increasing attention as a growing research area. Time-interval sequential patterns (also known as T-Patterns) not only capture the order of symbols but also the time delay between symbols, where the time delay is specified as a time-interval between a pair of symbols. Such patterns have been shown to be present in many different types of data (e.g. spike data, smart home activity, DNA sequences, human and animal behaviour analysis and the like) which cannot be captured by other pattern types. Recently, several mining algorithms have been proposed to mine such patterns from either transaction databases or static sequences of time-stamped events. However, they are not capable of online mining from streams of time-stamped events (i.e. event streams). An increasingly common form of data, event streams bring more challenges as they are often unsegmented and with unobtainable total size. In this paper, we propose a mining algorithm that discovers time-interval patterns online, from event streams and demonstrate its capability on a benchmark synthetic dataset.
AB - Temporal patterns that capture frequent time differences occurring between items in a sequence are gaining increasing attention as a growing research area. Time-interval sequential patterns (also known as T-Patterns) not only capture the order of symbols but also the time delay between symbols, where the time delay is specified as a time-interval between a pair of symbols. Such patterns have been shown to be present in many different types of data (e.g. spike data, smart home activity, DNA sequences, human and animal behaviour analysis and the like) which cannot be captured by other pattern types. Recently, several mining algorithms have been proposed to mine such patterns from either transaction databases or static sequences of time-stamped events. However, they are not capable of online mining from streams of time-stamped events (i.e. event streams). An increasingly common form of data, event streams bring more challenges as they are often unsegmented and with unobtainable total size. In this paper, we propose a mining algorithm that discovers time-interval patterns online, from event streams and demonstrate its capability on a benchmark synthetic dataset.
UR - http://www.scopus.com/inward/record.url?scp=85045843334&partnerID=8YFLogxK
U2 - 10.1007/978-981-13-0292-3_9
DO - 10.1007/978-981-13-0292-3_9
M3 - Conference paper
AN - SCOPUS:85045843334
SN - 9789811302916
T3 - Communications in Computer and Information Science
SP - 144
EP - 157
BT - Data Mining - 15th Australasian Conference, AusDM 2017, Revised Selected Papers
A2 - Stirling, David
A2 - Boo, Yee Ling
A2 - Chi, Lianhua
A2 - Ong, Kok-Leong
A2 - Liu, Lin
A2 - Williams, Graham
PB - Springer-Verlag Wien
T2 - 15th Australasian Conference on Data Mining, AusDM 2017
Y2 - 19 August 2017 through 20 August 2017
ER -