Research

Overview

In the Enterprise Systems Group we perform research and development on topics pertaining to information systems analysis and design, particularly model-driven business analysis.

We publish in highly-competitive international conferences and journals. A list of selected publications can be found here. Our RE’10, CAiSE’11 and ER’13 papers were all chosen to be amongst the top 5 of the conference or candidates for best paper awards.

We are currently working on three major research topics:

Decision Support for Systems Configuration
Decision Model Visualization Comprehension
Goal and Preference Acquisition

Decision Support for Systems Configuration

Off-the-shelf information systems, such as Enterprise Resource Planning (ERP) systems, are very popular amongst organizations interested in a reliable, mature and quickly implementable solution to common automation problems. Many of those systems are the result of decades of development and maintenance and have been adopted by a vast number of organizations, each posing their own structural and process particularities. To successfully accommodate larger and larger variations of business needs, however, such systems have inevitably grown very complex, requiring substantial amount of work to instantiate and fine-tune to the exact needs of the acquiring organization. As such, making and implementing configuration decisions can require substantial effort and experience.

Is there a way by which analysts can reduce the costs of off-the-shelf enterprise system configuration through automation? Advances in Artificial Intelligence (AI) and Conceptual Modeling tell us there could be. Our so far research with small and personal systems has shown that it is possible to reduce a large number of low-level configuration decisions into a smaller number of accessible high-level ones. Instead of manually researching and performing tedious configurations that also require specialized training, analysts may simply focus on developing models of intentional and process constraints and automatically compile them into the desired configuration. We are looking at the conceptual modeling and AI reasoning infrastructure as well as the decision support interface that is necessary for such automation to be effective.

Relevant Publications:
Saeideh Hamidi, Perilkis Andritsos and Sotirios Liaskos.  Constructing Adaptive Configuration Dialogs using Crowd Data.  In Proceedings of the 29th ACM/IEEE International Conference on Automated Software Engineering (ASE’14). Västerås, Sweden. pp 485-490. 2014.  (pdf) (ACM)
Sotirios Liaskos, Shakil M. Khan, Mikhail Soutchanski, John Mylopoulos: Modeling and Reasoning with Decision-Theoretic Goals. In Proceedings of the International Conference on Conceptual Modeling (ER 2013), 2013. (pdf) (bib) — the final publication is available at link.springer.com.
Sotirios Liaskos, Sheila McIlraith, Shirin Sohrabi, John Mylopoulos: Representing and reasoning about preferences in requirements engineering. Requirements Engineering Journal (REJ). 16(3). 2011. (pdf) (publisher) (bib).

top

Decision Model Visualization and Comprehension

Conceptual models (e.g. UML) enjoy an increasing attention from practitioners for describing various aspects of information systems. Models that represent decisions and their structures have also been proposed for use in various decision making activities within the information-systems analysis and design cycle. Such models are aimed at allowing decision makers comprehend the decision alternatives, the constituents of such alternatives, the relevant criteria and how all these components interact with each other to constitute a decision problem.
The way the model is presented, however, surely affects the decision. Is it a static representation or a dynamic? Is it diagrammatic, spatial or symbolic/textual? How is importance, relevance and preference represented? Focusing on decision problems pertinent to Information Systems, our experimental work focuses on understanding how visual or other features of the representation of a decision problem affects the decision maker’s comprehension of the problem as well as the quality of her decision.
Relevant Publications:
Sotirios Liaskos, Saeideh Hamidi, Rina Jalman: Qualitative vs. Quantitative Contribution Labels in Goal Models: Setting an Experimental Agenda. In Proceedings of the International i* Workshop (iStar’13). 2013. (pdf) (bib)
Sotirios Liaskos, Rina Jalman, Jorge Aranda: On eliciting contribution measures in goal models. In Proceedings of the 20th IEEE International Requirements Engineering Conference (RE’12). 2012. (pdf) (publisher) (bib)

top

Goal and Preference Acquisition

Goals now have a central position in business analysis. It is widely accepted that before we decide what functions the software system needs to have, we need to know what stakeholders want. It makes sense: neither do we want to go into the trouble of developing stuff that nobody wants, nor do we want to finish the project and go, having left out functions that people really wanted. An additional problem is priority and preference: when stakeholders have more than one goals, and these goals may be conflicting, an understanding of their relevant importance is necessary for conducting useful analysis.

But how do we learn what people want and prefer? Do we just ask them? Do they know what to say, i.e. do they really have well established goals and all they need is to tell us? And where do all these goals and preferences really come from? To answer these question we experimentally investigate how elicitation of goals and preferences is performed or should be realistically performed and what are factors that can distort or otherwise affect such elicitation. At the same time we look at novel interfaces by which elicitation can be conducted, such as natural language. Our goal is to both develop a descriptive theory of how elicitation is performed, and to provide tools to facilitate it.

Relevant Publications:
Fatima Alabdulkareem, Nick Cercone, and Sotirios Liaskos.  Goal and Preference Identification through Natural Language.  In Proceedings of the 23rd IEEE International Requirements Engineering Conference (RE’15). Ottawa, Canada. pp 56-65. 2015.  (pdf)
Sotirios Liaskos, Rina Jalman, Jorge Aranda: On eliciting contribution measures in goal models. In Proceedings of the 20th IEEE International Requirements Engineering Conference (RE’12). 2012. (pdf) (publisher) (bib)

top