TY - GEN
T1 - Preservation of integrity constraints by workflow
AU - Liu, Xi
AU - Su, Jianwen
AU - Yang, Jian
PY - 2011
Y1 - 2011
N2 - Integrity constraints on data are typically defined when workflow and business process models are developed. Keeping data consistent is vital for workflow execution. Traditionally, enforcing data integrity constraints is left for the underlying database system, while workflow system focuses primarily on performing tasks. This paper presents a new mechanism that turns a workflow into an equivalent one that will preserve integrity constraints. For a given workflow schema (or model) and a given set of data integrity constraints, an algorithm developed in this paper injects additional conditions into the workflow schema that restricts possible execution paths. The modified workflow will guarantee data consistency (i.e., satisfaction of the integrity constraints) whenever the workflow updates the database(s). In addition, we show that our injection mechanism is "conservative complete", i.e., the conditions inserted are weakest possible. By making workflow execution self-behaving, enforcing integrity constraints over multi-databases is avoided, and constraints specific to a workflow can also be enforced effectively. Mechanisms such as this enhance independence of workflow executions from the environment-a much desired property.
AB - Integrity constraints on data are typically defined when workflow and business process models are developed. Keeping data consistent is vital for workflow execution. Traditionally, enforcing data integrity constraints is left for the underlying database system, while workflow system focuses primarily on performing tasks. This paper presents a new mechanism that turns a workflow into an equivalent one that will preserve integrity constraints. For a given workflow schema (or model) and a given set of data integrity constraints, an algorithm developed in this paper injects additional conditions into the workflow schema that restricts possible execution paths. The modified workflow will guarantee data consistency (i.e., satisfaction of the integrity constraints) whenever the workflow updates the database(s). In addition, we show that our injection mechanism is "conservative complete", i.e., the conditions inserted are weakest possible. By making workflow execution self-behaving, enforcing integrity constraints over multi-databases is avoided, and constraints specific to a workflow can also be enforced effectively. Mechanisms such as this enhance independence of workflow executions from the environment-a much desired property.
UR - http://www.scopus.com/inward/record.url?scp=81255207796&partnerID=8YFLogxK
U2 - 10.1007/978-3-642-25109-2_6
DO - 10.1007/978-3-642-25109-2_6
M3 - Conference proceeding contribution
AN - SCOPUS:81255207796
SN - 9783642251085
VL - 7044
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 64
EP - 81
BT - On the Move to Meaningful Internet Systems, OTM 2011 - Confederated International Conferences: CoopIS, DOA-SVI, and ODBASE 2011, Proceedings, Part 1
A2 - Meersman, Robert
PB - Springer, Springer Nature
CY - Heidelberg
T2 - 10th Confederated International Conferences on On the Move to Meaningful Internet Systems, OTM 2011: CoopIS, DOA-SVI, and ODBASE 2011
Y2 - 17 October 2011 through 21 October 2011
ER -