TY - GEN
T1 - Local process model discovery
T2 - 39th International Conference on Application and Theory of Petri Nets and Concurrency, PETRI NETS 2018
AU - Tax, Niek
AU - Sidorova, Natalia
AU - van der Aalst, Wil M.P.
AU - Haakma, Reinder
PY - 2018/1/1
Y1 - 2018/1/1
N2 - This paper introduces the tool LocalProcessModelDiscovery, which is available as a package in the process mining toolkit ProM. LocalProcessModelDiscovery aims to discover local process models, i.e., frequent patterns extracted from event logs, where each frequent pattern is expressed in the form of a Petri net. Local process models can be positioned in-between process discovery and Petri net synthesis on the one hand, and sequential pattern mining on the other hand. Like pattern mining techniques, the LocalProcessModelDiscovery tool focuses on the extraction of a set of frequent patterns, in contrast to Petri net synthesis and process discovery techniques that aim to describe all behavior seen in an event log in the form of a single model. Like Petri net synthesis and process discovery techniques, the models discovered with LocalProcessModelDiscovery can express a diverse set of behavioral constructs. This contrasts sequential pattern mining techniques, which are limited to patterns that describe sequential orderings in the data and are unable to express loops, choices, and concurrency.
AB - This paper introduces the tool LocalProcessModelDiscovery, which is available as a package in the process mining toolkit ProM. LocalProcessModelDiscovery aims to discover local process models, i.e., frequent patterns extracted from event logs, where each frequent pattern is expressed in the form of a Petri net. Local process models can be positioned in-between process discovery and Petri net synthesis on the one hand, and sequential pattern mining on the other hand. Like pattern mining techniques, the LocalProcessModelDiscovery tool focuses on the extraction of a set of frequent patterns, in contrast to Petri net synthesis and process discovery techniques that aim to describe all behavior seen in an event log in the form of a single model. Like Petri net synthesis and process discovery techniques, the models discovered with LocalProcessModelDiscovery can express a diverse set of behavioral constructs. This contrasts sequential pattern mining techniques, which are limited to patterns that describe sequential orderings in the data and are unable to express loops, choices, and concurrency.
KW - Frequent pattern mining
KW - Petri nets
KW - Process discovery
UR - http://www.scopus.com/inward/record.url?scp=85048059331&partnerID=8YFLogxK
U2 - 10.1007/978-3-319-91268-4_20
DO - 10.1007/978-3-319-91268-4_20
M3 - Conference contribution
AN - SCOPUS:85048059331
SN - 978-3-319-91267-7
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 374
EP - 384
BT - Application and Theory of Petri Nets and Concurrency
A2 - Khohamenko, V.
A2 - Roux, O.H.
PB - Springer
CY - Dordrecht
Y2 - 24 June 2018 through 29 June 2018
ER -