Servicenavigation

Project Details

Back

Projekt Print View

Negotiations: A Model for Tractable Concurrency.

Subject Area Theoretical Computer Science
Term from 2015 to 2019
Project identifier Deutsche Forschungsgemeinschaft (DFG) - Project number 273811150
 
Final Report Year 2020

Final Report Abstract

In business processes different agents communicate and conduct actions together in order to complete a task. For example, diagnosing and treating a patient is a process involving the patient, his or her general practitioner (GP), possibly a specialist, the receptionist at the hospital, doctors and nurses, a radiologist, etc. The process starts with an action, say “make an appointment with the GP”, jointly executed by the patient and the GP, and ends with an action, say “discharge from hospital”, with a doctor and the patient as agents. The actions in-between must need to be executed in sequence, but may also be alternatives (e.g., conduct test A or test B) or concurrent (e.g., conduct tests A and B in parallel). Business processes can be complex, and when enterprises design them they make errors, or design them inefficiently. Analyzing complex business processes requires computer support. For that, one needs description formalisms for the representation of business processes, and efficient algorithms for their automatic analysis. Negotiation diagrams are a formalism for the description and analysis of business processes introduced by the principal investigator and its co-authors shortly before the start of the project. They are very close to another formalism called workflow Petri nets, very popular in industry, but less adequate for theoretical investigations. The project has developed many algorithms for the automatic analysis of negotiation diagrams and workflow Petri nets. In particular, it has studied so-called deterministic negotiation diagrams, in which the next action in which an agent has to participate is completely determined by the result of the last action in which it participated. The algorithms are orders of magnitude faster than previous ones, and allow one to analyze many qualitative and quantitative questions, like the possibility to reach a deadlock (a situation in which all agents are waiting for other agents to conduct an action), and quantitative ones, like the expected cost or the expected time of executing a given business process. The algorithms design during the project have been partly implemented in the framework ProM, a popular tool for analyzing business processes.

Publications

 
 

Additional Information

© 2024 DFG Disclaimer / Copyright / Privacy Policy

Textvergrößerung und Kontrastanpassung