Table of Contents

More efficient, lower-cost internet services through research by Mathematics and Computer Science department

The Artifact-Centric Service Interoperation (ACSI) project started in June 2010. This European research project aims to enable faster, smoother internet services to be provided by developing new techniques and tools. With the help of process mining techniques, four researchers in the Architecture of Information Systems group (Department of Mathematics and Computer Science) will analyze the new systems and tools on which the project focuses.

ACSI is a partnership between the Architecture of Information Systems research group led by Wil van der Aalst (TU/e) and IBM, Sapienza Universita degli Studi di Roma, Free University of Bozen-Bolzano, Imperial College, University of Tartu, Indra Software Labs and Collibra. The project is led by IBM Research in Haifa, Israel.

Linked e-services

“Many e-services are linked together”, explains prof.dr.ir. Wil van der Aalst. “If you book a flight ticket online, you’re dealing with a reseller who in turn has to coordinate what he offers with airlines and insurance companies, for example. The coordination between all those parties is a very complex problem. And if there’s an error in any of the systems concerned, it’s easy for a message to be lost somewhere, with all the possible miscommunication and other problems that can lead to.”

Common platform

The aim of ACSI is to analyze coordination problems of this kind, and where possible to prevent them. Van der Aalst mentions ‘Easy Chair’, a system for reviewing papers which is used worldwide. He compares it to a whiteboard or a notepad, on which all the involved parties can write down what they want to say. “That means everyone concerned has the same, shared picture of the current situation, which eliminates the need for all kinds of bilateral links between many different parties. That’s the kind of common platform that ACSI is focusing on.”

While many current systems are either data- or process-oriented, according to Van der Aalst, ACSI aims to bring the two approaches in balance. The researchers also want to apply new forms of analysis, which is the core of the TU/e contribution. “If you use this kind of common platform, huge amounts of data are recorded. Using process mining, we are trying to accurately identify how people use a system, and where the bottlenecks are located.” Van der Aalst says that the software which his group has developed for this purpose, one of the central systems used within ACSI, is already being used by more than 100 companies. The total project has a duration of three years and a budget of 4.6 million euro.