Computer Science Colloquium


23. June 2011

Conjunctive Query Entailment in Description Logics with Counting, Inverses and Nominals

Dr. Sebastian Rudolf
Institute AIFB, Karlsruhe Institute of Technology (KIT)

Conjunctive queries (CQs), the standard query language in databases, have recently gained significant attention for querying DL knowledge bases. Several different techniques are available for a wide range of DLs. Nevertheless, for OWL DL and OWL 2, decidability of CQ entailment is an open problem. So far, the combination of nominals, inverse roles, and number restrictions caused unsolvable problems. We tackle this problem and present a decidability result for entailment of unions of CQs in a DL with all three problematic constructors. For queries with only simple roles, our result also shows decidability in the logic that underpins OWL DL.

Bio:

Sebastian Rudolph is senior researcher and project leader at the Institute AIFB, Karlsruhe Institute of Technology (KIT). He obtained his PhD in mathematics at the Institute for Algebra at the Dresden University of Technology in 2006. His active research interests include algebra, complexity theory, logic, machine learning, optimization, database theory, and computational linguistics. Sebastian coauthored two textbooks on the foundations of the Semantic Web. He is a guest lecturer at Heidelberg University, guest researcher at FZI Karlsruhe, and he was a member of the W3C OWL Working Group. As a guest researcher, he visited Oxford University, the Technical University of Vienna and the Laboratoire d′Informatique, de Robotique et de Microélectronique de Montpellier.


13. April 2010

Information Flow Control with Program Dependence Graphs

Jens Krinke
CREST, King's College London

In classic information flow control (IFC), noninterference guarantees that no information flows from secret input channels to public output channels. However, this notion turned out to be overly restrictive as many intuitively secure programs do allow some release, and thus intransitive noninterference allows declassification. We developed a static analysis that allows intransitive noninterference based on Program Dependence Graphs. In contrast to type systems that annotate variables, our approach annotates information sources and sinks. Our approach is flow- and context-sensitive and because it is based on an industrial strength Program Dependence Generator (CodeSurfer from GrammaTech), it allows IFC for realistic C programs.

Bio:

Jens Krinke is lecturer in the Department of Computer Science, King's College London, and Deputy Director of CREST - the Centre for Research on Evolution, Search and Testing. His current research topics include program analysis for software engineering purposes, in particular for bug detection, taint analysis and information flow control. Currently, he also works on Clone Detection and Code Provenance.


10. December 2009

The Past, Present and Future of Software Evolution

Time: Thursday 4pm, Location: SBS95, E4.040

Michael W. Godfrey
University of Waterloo, Canada

As Fred Brooks famously noted, change is an essential characteristic of software development, as software systems must respond to evolving requirements, platforms, and other environmental pressures. In this talk, we discuss the concept of software evolution from several perspectives. We examine how it relates to and differs from software maintenance. We discuss insights about software evolution arising from Lehman’s laws of software evolution and the staged lifecycle model of Bennett and Rajlich. We compare software evolution to other kinds of evolution, from science and social sciences, and we examine the forces that shape change. Finally, we discuss the changing nature of software in general as it relates to evolution, and we propose open challenges and future directions for software evolution research.

This is joint work with Prof Daniel German of the University of Victoria, and is based on an invited paper for the Frontiers of Software Maintenance track at the 2008 IEEE International Conference on Software Maintenance.

Bio:

Michael W. Godfrey is an Associate Professor in the David R. Cheriton School of Computer Science at the University of Waterloo (Canada), where he is also a member of SWAG, the Software Architecture Group. Between 2001 and 2006, he held an Associate Industrial Research Chair in telecommunications software engineering sponsored by Nortel Networks and NSERC. He holds a PhD in Computer Science from the University of Toronto (1997) and was previously a faculty member at Cornell University. His main research area is software evolution: understanding how and why software changes over time. His research interests include evidence-based software engineering, software clone analysis, software tool design, reverse engineering, and program comprehension.


1. December 2009

Semantische Integration durch Reinterpretation - Ein Formales Modell

Özgür Özcep
Universität Hamburg

Die Integration von Wissen aus heterogenen Wissensquellen erfordert u.a. die Behandlung von terminologischen Konflikten, die durch die Ambiguität von Symbolen erklärt werden können. Im Vortrag wird ein Ansatz zur semantischen Integration bzgl. eines Integrationsszenarios vorgestellt, in dem die zu integrierenden Wissensquellen gut entwickelte und verwandte Ontologien mit potentiellen Ambiguitätskonflikten sind. Neben der Darstellung von postulatsorientierten Adäquatheitskriterien für dieses Integrationsszenario werden im Vortrag konkrete Integrationsoperatoren beschrieben, die auf dem Prinzip der Reinterpretation beruhen und die auf Techniken der Belief-Revision zurückgreifen. Die Gemeinsamkeiten und Unterschiede zwischen klassischen Belief-Revisionsfunktionen und Reinterpretationsoperatoren werden anhand von erfüllten Postulaten und anhand der Konstruktionsprinzipien erläutert.


24. November 2009

Qualitative Spatial and Temporal Reasoning: From Theory to Applications
-- The SparQ Toolbox --

Diedrich Wolter
Universität Bremen

Qualitative Spatial and Temporal Reasoning (QSTR) is the subfield of knowledge representation and symbolic reasoning that is involved with infinite spatial and temporal domains. Qualitative representations only capture relevant relations and abstract from unnecessary details. They are especially suited to handle uncertain or incomplete information. Exploiting the domain structure makes QSTR effective and efficient. One particular goal of QSTR is to reach for formal methods capable of human-level spatial and temporal reasoing. Thus, QSTR is relevant for human-machine interfaces, but QSTR can provide effective means for a variety of other applications as well. Within this talk, we review the essential QSTR reasoning techniques and evaluate how the state-of-the-art suits application requirements. We argue for the need of reference implementations and present the toolbox SparQ developed in our project. With our research and the development of SparQ we aim for making applications benefit from QSTR. Dr. Diedrich Wolter is principal investigator of the project qualitative reasoning about paths, shapes, and configurations at the University Bremen in the framework of the interdisciplinary Transregional Collaborative Research Center Spatial Cognition: Reasoning, Action, Interaction. The project develops qualitative representations and reasoning methods for high-level agent control.


9. November 2009

IBM Faculty Awards

Erwin Jung, IBM University Relations Manager Deutschland

The IBM Ph.D. Fellowship Program is an intensely competitive worldwide program, which honors exceptional Ph.D. students who have an interest in solving problems of interest to IBM and which are fundamental to innovation in many academic disciplines and areas of study. These include: computer science and engineering, electrical and mechanical engineering , physical sciences (including chemistry, material sciences, and physics), mathematical sciences (including optimization), business sciences (including financial services, communication, and learning/knowledge), and service sciences, management, and engineering (SSME).

Download slides


21. Oktober 2009

Software Engineering in Practice: How to realize a software project as large as Windows 7

Jon DeVaan, Senior Vice President, Windows Core Operating System Division, Microsoft Corporation


Vizepräsident Prof. Dr. Garabed Antranikian
und Microsoft-Vice President Jon DeVaan.

Einblicke in die Entwicklung des Windows-Betriebssystems haben alle erhalten, die am Mittwoch, 21. Oktober, an der TUHH den Vortrag des Senior Vice President von Microsoft, Jon DeVaan, verfolgten. Der Chefentwickler, ein studierter Mathematiker und Informatiker, erläuterte vor einer interessierten Zuhörerschaft im Raum 2022 die Kunst des Projektmanagements am Beispiel der Entwicklung von Windows 7. Etwa 1000 Microsoft-Mitarbeiter, hauptsächlich in den USA und Indien, sind an der Entwicklung dieses komplexen Systems beteiligt gewesen, die in drei Phasen verläuft, die sich solange wiederholen, bis das Produkt fertig ist: In der ersten Stufe arbeiten Gruppen bis zu 40 Mitarbeitern an den einzelnen Säulen der zukünftigen Software, danach findet in der zweiten Phase die Qualitätsprüfung bei der Integration der Säulen statt, und am Ende steht das Softwareprodukt. Anschließend werden diese Phasen wiederholt udn dadurch das Produkt immer weiter verfeinert. Plan the work und work the plan!, sei das Leitmotto, sagte DeVaan. Dabei wurde auch betont, welche Bedeutung Microsoft heute dem Thema Sicherheit und dem Security Development Lifecycle beimisst.

Download slides


29. September 2009

Visualisierungstechniken für Ergebnisse der Stochastischen Simulation

Dietmar Vogt

Computer-Aided Engineering und insbesondere numerische Simulationen sind ein wesentlicher Bestandteil der modernen Produktentwicklung. Für Produkte, die ein hohes Maß an Zuverlässigkeit und Robustheit erfordern, ist die Vorhersage der Auswirkungen produktions- und einsatzbedingter Unsicherheiten von großer Bedeutung. Zu diesem Zweck wird in wachsendem Maße Stochastische Simulation eingesetzt. Die Auswertung der umfangreichen stochastischen Simulationsergebnisse stellt den Ingenieur jedoch vor ein methodisches Problem, da die verfügbaren Methoden der statistischen Datenvisualisierung und der wissenschaftlichen Visualisierung nur begrenzt geeignet sind.

Im Rahmen des Vortrags werden daher neuartige Methoden zur Visualisierung und systematischen Analyse stochastischer Simulationsergebnisse von zeitlich und geometrisch diskretisierten Modellen aufgezeigt. Anschließend wird das Potential der Visualisierungsmethoden an exemplarischen Anwendungsfällen demonstriert.


5. August 2009

How to Handle Ontologies

Alex Borgida, Volker Haarslev, Ralf Möller

Conceptual data models play a major role in Computer Science as large-scale agent systems are designed such that computational entities can migrate over the web, can interact, and can contact world-wide information repositories. In these contexts a shared understanding between agents based on formalized conceptual data models (ontologies) is mandatory. In particular, it must be accepted that none of the agents has (or can obtain) complete information about the world. Thus, query answering with respect to ontologies and incomplete information becomes more and more important. In addition, modules have to be defined such that information can be sensibly shared and reasoned upon. Many research activities in the world manifest the importance of this research field, and joint acitivities have to be started to further clarify the relation between advanced database systems and practical knowledge representation and reasoning systems.

The Institute for Software Systems (STS) at Hamburg University of Technology invites two leading experts in this field, Alex Borgida (USA) and Volker Haarslev (Canada). They will give presentations about parts of their work in the above-mentioned research area. The two presentations will be complemented with a presentation about our work at STS. Talks will be given in a special colloquium at STS, with which we would like to encourage research collaborations in Computer Science between Hamburg University of Technology (as well as other universities in Hamburg) and well-known institutes and groups in the world.

The presentations are described below:


What, if anything, does it mean to Import Knowledge

Alexander Borgida
Department of Computer Science, Rutgers University, New Brunswick, NJ, USA

It is argued that, given an ontology (or a program for that matter), there is a difference between partitioning it into 'modules' and importing information about some set of identifiers defined in it. We start by surveying a sample of lesser known papers on this topic in both the AI and Database literature. Based on this, we derive a list of desirable properties for the notion of 'importing a set of identifiers from an exporting to an importing knowledge base', and provide a framework for alternate formal definitions of this notion. The best-known proposal starts from the logical notion of 'conservative extension' [Cuenca Grau et al] but is dependent on the syntactic formulation of the ontology. Our proposal suggests conceptually expanding the ontology ( adding trivial inferences, making it more 'fine-grained' [Horridge et al], etc.) and allowing limitations on the use of imported terms. We also discuss the difficulties of minimizing the number of concepts rather than the number of axioms during importing.

Download slides


Query Answering with Ontologies - Optimizing in EXPTIME and beyond: lessons learnt and challenges ahead

Volker Haarslev
Department of Computer Science, Concordia University, Montreal, Quebec, Canada

In this talk I will briefly introduce description logics / OWL-DL and associated inferences services. Afterward I will discuss the architecture of the description logic reasoner Racer and highlight selected tableau optimization techniques, especially on assertional reasoning. Several recently devised optimization techniques were introduced due to requirements from semantic web applications relating huge amounts of (incomplete) data to ontological information.

Download slides


Querying Large-Scale Ontologies

Ralf Moeller
Hamburg University of Technology, Software Technology and Systems Institute (STS), Hamburg, Germany

Terabytes of data will be a common size to be managed on personal computers in the not too far future, and database technology has matured in such a way that this amount of data is manageable to a large extent even if queries are defined in an ad hoc manner. In particular, a common assumption in database applications is that the conceptual schema is used for deriving the implementation schema (e.g., relational schema), which is then used for storing data. Views might relate notions from the conceptual to the implementation schema, and if updates are neglected, views can easily be used to manage a database with respect to the conceptual schema. In almost all practical database systems, data is considered to be complete, with a corresponding impact on query answering. In new application contexts such as the semantic web, however, software agents migrate to new sites, and they use their conceptual schema for querying large data repositories found at different sites. With new schemata being used, data can hardly be seen as complete, and thus, query answering w.r.t. incomplete information becomes more and more important. Using ontologies, query answering with respect to views and incomplete data descriptions becomes possible. In the tutorial we present recent advances in query answering techniques for for large sets of data descriptions w.r.t. large and expressive ontologies (large sets of axioms specified in an expressive language) under the assumption that data descriptions are assumed to be incomplete. We present query languages which can be practically used in combination with ontologies of varying expressiveness.

Download slides


Pictures showing some members of the audience on Aug. 5th:

We would like to thank Joachim Schmidt and Resa Rasouli for providing pictures of the event.

If somebody shown on these pictures would like to have the corresponding pictures removed, please send an email to r.f.moeller --at-- tu-harburg.de.


13.7. 2009

Understanding Agent-Based Models of Barter Economies through Specifications

Nicola Botta
Potsdam Institute for Climate Impact Research

In applied mathematics, physics and engineering, computer-based models are traditionally developed and applied to solve well-defined *problems*. These problems are often expressed through equations. The *terms* of the problems -- what is given and what is sought -- are usually stated explicitly or they are unambiguously implied by the context. Often, such problems have been extensively studied and theoretical results, e.g., about existence of solutions and their stability, are available. This knowledge provides quality measures for computer-based models: developers, implementors and practitioners have meaningful ways to assess the quality of implementations, compare, improve and extend models without resorting to (often unavailable or poor) empirical data. In contrast, agent-based models of socio-economic systems are often developed through *exploratory programming*. They are not designed to solve well-defined problems and are usually subject to weak accountability requirements (with the exception, maybe, of models for financial markets). Often, agent-based models of socio-economic systems are used a workbenches to investigate conjectures and gain insights about the behavior of ``real'' economic systems. For these purposes, exploratory programming is, on the short run, fast, flexible and cheap. However, exploratory programming is error-prone and provides little support for model validation, assessment and inter-comparison. The absence of high-level problem descriptions is a major obstacle to model understanding, communication and sharing: it forces scientists to understand models through time-consuming low-level code reading or through the free interpretation of narrative model descriptions. In my talk I present examples of agent-based model descriptions at an intermediate level. This level is more abstract than code and more precise than narrative descriptions. It is based on standard mathematical notation and on elements of the functional programming language Haskell. I argue that such a *specification* level effectively supports scientists in understanding and in communicating models. It provides useful guidelines for model implementation and validation and helps making model extension and re-factoring both faster and safer.