Keynote Lectures

Abstract

In today’s era of data sharing, immediate communication and world-wide distribution of participants, at a time when teams are asked to be ever more agile, the traditional approach of model repositories no longer meets expectations. Centralized organization has become inconsistent with the way in which the world and its companies function.

In today’s world, it is virtually impossible to set up a model repository for different enterprise entities, large-scale systems or projects, which can be accessed by all participants (readers, contributors, partners, and so on). Standard techniques based on a centralized repository with a designated manager come up against a vast variety of situations, with participants who neither want nor are able to conform to uniform rules and management.

This does not allow model-based knowledge management at an enterprise or global level. It inhibits agility and open team cooperation. We believe that this is a major hurdle to the dissemination of model-based approaches; the reality of heavyweight model management hinders the most appealing of model-based approaches.

Based on the latest technologies and research for model repositories, this talk will explain why current model repository technologies are a major drawback and will present a way of supporting highly decentralized organizations, and agile and open team cooperation. Scaling up and widening the scope of model repositories will enable modeling support to be applied to the "extended enterprise", which incorporates its eco-system (providers, partners, and so on).

Abstract
Given the fundamental role of domain models in software and knowledge engineering, it is important that they be as precise and detailed as possible whilst at the same being simple and easy-to-understand. This presents a challenge, however, because higher precision and detail often lead to more complexity and obscurity. To overcome this tension it is necessary to reconcile approaches focussed on maximizing the richness of domain models with approaches focussed on maximizing their simplicity. The former (richness) is addressed by vocabularies (i.e. sets of linguistic modeling concept (a.k.a. metamodels)) optimized to capture the full range of real world phenomena perceived by human beings. This is the role of foundational (a.k.a. upper level) ontologies such as BWW or UFO. The latter (simplicity) is achieved by modelling infrastructures that allows the basic ingredients of conceptual modelling (classification and generalization) to be expressed with minimum accidental complexity. This is the focus of deep (i.e. multi-level) modelling. In this talk Colin Atkinson will discuss the challenges faced in reconciling the goals of richness and simplicity in conceptual modelling and the benefits that can be gained through a synergetic integration of foundational ontologies and deep modelling.

Abstract

Product architecture derivation is a crucial activity in Software Product Line (SPL) development since an inadequate decision during the architecture design directly impacts the quality of the product under development. Deriving individual products from shared software assets is a time-consuming and expensive activity. Although some methods for architecture derivation and evaluation have been proposed over the past years, there are a number of challenging issues as to how we derive product architectures that meet the required quality attributes for the system. In this talk, I will overview the state-of-the-art in using models for architecture derivation and evaluation, will discuss the challenges faced in deriving and evaluating product architectures and the benefits that can be gained from the use of a multimodel (i.e., a set of interrelated models that represents different viewpoints of a particular system) that allows the software engineer to analyze the impact of different design choices or software architectures before they are reified into runnable code.

Abstract
Timed automata, priced timed automata and energy automata have emerged as useful formalisms for modeling a real-time and energy-aware systems as found in several embedded and cyber-physical systems. Whereas the real-time model checker UPPAAL allows for efficient verification of hard timing constraints of timed automata, model checking of priced timed automata and energy automata are in general undecidable -- notable exception being cost-optimal reachability for priced timed automata as supported by the branch UPPAAL Cora. These obstacles are overcome by UPPAAL-SMC, the new highly scalable engine of UPPAAL, which supports (distributed) statistical model checking of stochastic hybrid systems with respect to weighted metric temporal logic. The talk will review UPPAAL-SMC and its applications to the domains of energy-harvesting wireless sensor networks, schedulability and execution time analysis for mixed criticality systems, battery-aware scheduling with respect to correctness, fault- and performance analysis. In the talk I will also indicate how UPPAAL SMC may play be of benefit for counter example generation, refinement checking, testing, controller synthesis and optimization.

Abstract
One of the grand challenges in our networked world are the large, weakly structured and unstructured data sets. This is most evident in Biomedicine (Medical Informatics + Bioinformatics): The trend towards personalized medicine results in increasingly large amounts of (-omics) data. In the life sciences domain, most data models are characterized by complexity, which makes manual analysis very time-consuming and often practically impossible. To deal with such data, solutions from the machine learning community are indispensable and it is marvelous what sophisticated algorithms can do within high-dimensional spaces. We want to enable a domain expert end-user to interactively deal with these algorithms and data, so to enable novel discoveries and previously unknown insights. Our quest is to make such approaches interactive, hence to enable a computationally non-expert to gain insight into the data, yet to find a starting point: “What is interesting?”. When mapping the results back from arbitrarily high-dimensional spaces R^n into R^2 there is always the danger of modeling artifacts, which may be interpreted wrongly. A synergistic combination of methodologies and approaches of two areas offer ideal conditions towards working on solutions for such problems: Human-Computer Interaction (HCI) and Knowledge Discovery/Data Mining (KDD), with the goal of supporting human intelligence with machine intelligence. Both fields have many unexplored, complementing intersections and the aim is to combine the strengths of automatic, computer-based methods, both in time and space, with the strengths of human perception and cognition, e.g. in discovering patterns, trends, similarities, anomalies etc. in data.