Colloquium

If not stated otherwise, seminar usually takes place on Tuesday at 2pm in room S8


Future colloquiums



The Data Replication Problem for Flexible Query Answering

Date Speaker Affiliation
2.12.2014 Lena Wiese Institut für Informatik, Fakultät für Mathematik und Informatik, Georg-August-Universität Göttingen
Abstract:

The data replication problem (DRP) describes the task of distributing copies of data records (that is, database fragments) among a set of servers in a distributed database system. For the application of flexible query answering, several fragments can be overlapping (in terms of rows in a database table). The optimization goals of the DRP are such that fragments are assigned to servers without exceeding their capacity and – for a fixed replication factor m – each table row can be found at at least m servers. This formulation of the DRP is an extension of the Bin Packing Problem with Conflicts (BPC) where the conflict graph must be colored with a defective coloring.

Benchmarking encrypted Data Storage in HBase and Cassandra

Date Speaker Affiliation
2.12.2014 Tim Waage Institut für Informatik, Fakultät für Mathematik und Informatik, Georg-August-Universität Göttingen
Abstract:

A major disadvantage of cloud storage providers is their lack of security features. We benchmarked and quantified the performance loss that such cloud storage systems incur due to encrypted storage, in particular the NoSQL databases Apache Cassandra and HBase. Therefore we did two tests. On the one hand we used the YCSB (Yahoo! Cloud Serving Benchmark), which we extended to add encryption/decryption logic. On the other hand we implemented a practical scheme for searching over encrypted text, which we benchmarked with real world datasets and Cassandra.


Past colloquiums



Spectral clustering methods for classifying words from different languages in plutilingual corpora

Date Speaker Affiliation
20.05.2014 Pascal Vaillant Services et réseaux de communication, IUT de Bobigny, Université Paris-13 (Paris-Nord), France
Abstract:

Our work is addressing the problem of trying to deal with plurilingual corpora, that is natural language corpora where speakers spontaneously use different languages and mix them to some degree. A useful preliminary task is then to draw the borders between the different languages used in the text; the difficulty is that those borders often are fuzzy. The information about lexical units belonging to one language or another is implicitly contained in the texts themselves, due to the fact that words from a same language tend to appear close to another, in the same segments of text. We have explored the question of how automatic learning algorithms may be used to gather this implicit information from non-annotated, or partially annotated, corpora. We project the linguistic units observed in the text onto a high dimensional space where they tend to appear clustered together. Spectral analysis methods are used to define the most relevant clusters. When a corpus is made up of a concatenation of linguistically homogeneous segments, the categories which spontaneaously come out of the clustering are related to the different languages used.

Algorithmic Trading and Experimental Computational Simulation Environments

Date Speaker Affiliation
19.11.2013 Michal Galas University College London, UK
Abstract:

Speciálně vhodné pro zájemce o:

  • Big & Linked Data
  • Knowledge extraction
  • Parallel processing
  • Algoritmic trading
  • Spolupráci s prestižní britskou univerzitou obecně

Efficient Query Processing in Probabilistic-Temporal Databases

Date Speaker Affiliation
11.06.2013 Martin Theobald University of Antwerp, Antwerp, Belgium
Abstract:

Recent advances in the field of information extraction have paved the way for the automatic construction and growth of large, semantic knowledge bases from Web sources. Knowledge bases like DBpedia or YAGO today contain hundreds of millions of facts about real-world entities and their relationships among each other, which are captured in the popular Resource Description Framework (RDF) format. However, the very nature of the underlying extraction techniques entails that the resulting RDF knowledge bases may face a significant amount of incorrect, incomplete, or even inconsistent factual knowledge, which makes efficient and reliable query answering over this kind of uncertain RDF data a challenge. Our query engine, coined URDF, performs query answering in uncertain RDF knowledge bases via a combination of Datalog-style deduction rules, consistency constraints, and probabilistic inference, which will be the main subject of this talk. Specifically, by casting the above scenario into a probabilistic database setting, we develop a new top-k algorithm for query answering, which - for the first time in the context of probabilistic databases - allows us to fully integrate data and confidence computations over this kind of probabilistic input data. Extensions of our framework include the automatic learning of these deduction rules from RDF data sources, as well as the consideration of temporal deduction rules and consistency constraints over time-annotated, probabilistic facts.

Matrix Factorization Techniques for Recommender Systems [slides]

Date Speaker Affiliation
27.11.2012 Tomáš Horváth Institute of Computer Science, Faculty of Science, Pavol Jozef Šafárik University, Košice, Slovakia
Abstract:

Recommender Systems (RS) [1, 2] have became very important part of many e-Commerce applications. The aim of a RS is to recommend items to users they are more likely to be interested in. Recommendation is based on a so-called user model learned from the available data such as any type of domain knowledge, user/item attributes, the recorded user-item interactions, etc. A user model can be personalized or non-personalized from which the former has got more atten- tion. The tutorial provides an overview of the main concepts, techniques and directions of recommender systems, a relatively new research area lying on the border of information retrieval, data mining and computer-human interaction. In the frst part of the tutorial, basic concepts are introduced followed by the defnition of the general recommendation task. Depending on the type of user feedback, we distinguish between rating prediction and item recommendation. In the second part, the collaborative-filtering technique is described with focus on matrix factorization (MF) [3] since it is the most popular recommendation technique. In the third part of the tutorial, two well-known issues, the cold-start problem and context-aware recommendation are discussed followed by evaluation issues. At the end, the MyMediaLite recommender systems library [4], a state-of-the art freeware implementation of major recommender techniques, is described. Acknowledgements I would like to thank to my colleagues at the Infor- mation Systems and Machine Learning Lab of the University of Hildesheim, Germany as well as to Štefan Pero for their advices and help in preparing this tutorial which was also partially supported by the grant VEGA 1/0832/12.

References:
  • 1. D. Jannach, M. Zanker, A. Felfernig, G. Friedrich. Recommender Systems: An In- troduction. Cambridge University Press, 2010.
  • 2. F. Ricci, L. Rokach, B. Shapira, P.B. Kantor. Recommender Systems Handbook. Springer, 2011.
  • 3. Y. Koren, R. Bell and C. Volinsky. Matrix Factorization Techniques for Recom- mender Systems. COMPUTER (Vol. 42, 8): IEEE Computer Society Press, Los Alamitos, USA, 2009.
  • 4. Z. Gantner, S. Rendle, C. Freudenthaler, L. Schmidt-Thieme. MyMediaLite: A Free Recommender System Library. In the Proceedings of the 5th ACM International Conference on Recommender Systems, Chicago, USA, 2011.

Modelovanie softvérovej architektúry pomocou architektonických pohľadov [slides]

Date Speaker Affiliation
30.10.2012 Ľubor Šešera SOFTEC & FIIT STU, Bratislava
Abstract:
  • 1. Čo je to softvérová architektúra?
  • 1.a. Genéza pojmu softvérová architektúra: od Parnasa k Shaw
  • 1.b. Vyčlenenie architektúry ako samostatnej oblasti softvérového inžinierstva Perrym a Wolfom
  • 1.c. Definícia softvérovej architektúry podľa štandardu ISO 42010
  • 2. Prehľad prístupov k architektonickým pohľadom:
  • 2.a. Pohľady v metódach analýzy a návrhu: od Constantina k UML
  • 2.b. Pohľady v softvérovej architektúre: Kruchten, Rozanski a Woods, Clements a kol.
  • 2.c. Generalizačné prístupy: Zachman, TOGAF a ArchiMate, štandard ISO 42010
  • 3. Modelovanie architektúry bankového systému použitím pohľadov Rozanského a Woodsa a jazyka UML

An implicit approach to deal with periodically repeated data [slides]

Date Speaker Affiliation
9.10.2012 Bela Stantic Griffith University, Australia
Abstract:

Periodic data play a major role in many application domains, spanning from manufacturing to office automation and from scheduling to data broadcasting. In many of such domains, the huge number of repetitions makes the goal of explicitly storing and accessing such data very challenging. In this talk a new methodology based on an implicit representation of periodic data will be explained. It will be shown that the proposed model captures the notion of periodic granularity, provided by the temporal database glossary, and is an extension of the TSQL2 temporal relational data model. At first, the algebraic operators will be defined and then access algorithms will be introduced. Finally, results from extensive experimental evaluation will be presented, which demonstrate that the implicit representation of periodic data outperforms the explicit approach.

Dr Bela Stantic is member of Institute for Integrated and Intelligent systems within the Griffith University, Brisbane, Australia, and currently visiting Professor at University of Novi Sad. His area of Research are management of complex data structures. He has published more than 70 conference, Journals and books chapters. He was/is member of many Program Committees in area of his research and was/is editor for many Journals.

Opinion Mining from Social Media [slides]

Date Speaker Affiliation
25.09.2012 Dr. Diana Maynard Natural Language Processing Group, University of Sheffield

Rozvoj IT - jízda na tygru cs

Date Speaker Affiliation
13.3.2012 Jaroslav Král KSI MFF UK
Abstract:

Heroické příběhy prvých počítačů a také příklady lidské hlouposti při začátcích IT v Čechách. Problém opomíjení triviálních skutečností – problém formulace správné otázky. Co jsme si tehdy nedovedli představit v dobrém (růst výkonu a ovlivňování každodenního života) a co ve zlém. Proklaté dědictví ASŘ(automatizovaných systémů řízení). Stále nechápeme klíčové přínosy IT (IT je v mnoha statistikách kromě těch, které dokládají přínosy) a hlavní hrozby (úřadů je stále více a staví si stále větší paláce, proč nelze pracovat doma). Hrozby vodopádu a velkého třesku v IT i jinde. Případ Business Process Restructuring a jeho častý skrytý účel. IT systémy "založené na znalostech", které ničí klíčové existující znalosti a dovednosti. Zlepšuje IT podstatně management (výběr manažerů, krize a chybějící snaha o udržitelnost)? Asi ne, ale kritiku tohoto stavu necháváme na umělci. Opomíjené důsledky toho, že SW produkty jsou technické artefakty se vším všudy, jejichž tvorba vyžaduje určitou dobu (viz Drábkův systém) a vedle IT znalostí stále více sociální a víceoborové dovednosti (nestačí o tom jen něco přečíst, byť z dobré učebnice) a existenci legislativy, která má hlavu a patu a není cinknutá.

Legislativa proti kvalitě dat.

Evidentní aspekty SW inženýrství platné i mimo IT:

  • důsledky přebíráni modelu vzdělávání od velkých států a jejich korelace s důsledky změny velikosti uživatele informačních systému,
  • zlepšování kvality dat (dostupnost, copyright) a ochrana všech lidských práv a nikoliv jen některých, proč zákonodárci nedělají analýzu rizik a raději přikazují zhoršovat kvalitu dat,
  • meze možností evaluace kvality vzdělání a zlepšování zpětných vazeb ve společnosti prostřednictvím IT,
  • meze elektronizace výuky a automatizace byznys procesů, a jak obojí souvisí s tím, proč se nedaří pracovat doma,
  • hrozby pro naše absolventy plynoucí z přílišné specializace na činnosti, které je obtížné dělat ve vyšším věku – už po čtyřicítce, aneb jak nebýt na podpoře od pětatřiceti,
  • odrovná ajťáky /open space/ pracovní prostředí?
  • pravidla podpory výzkumu jako zbraň proti víceoborovým znalostem a aplikacím výsledků výzkumu, IT dělá vědců účtaře amatéry
  • proč IT neumožnily predikci krize z roku 2008 na rozdíl od Rusa, který ji, spolu s velkou krizí z roku 1930, předpověděl v roce 1924,
  • asociálnost sociálního IT,
  • některé dramaticky pesimistické názory na IT ve statni správě (Tainter)

IT je také řemeslo, které má zatím pro naše dovedné studenty zlaté dno. Jde o to, aby to nebylo jen na pár let a víko jejich řemesla nebylo příliš podělané.

IT mění svět. Existují ale silné hrozby. Mnohé z nich zatím je jen tušíme, mnohé už dávno nejsou za obzorem. Ale s tohohle stále rychleji pádícího tygra nemůžeme seskočit, my ajťáci i ti ostatní.

Preference SQL - An Overview

Date Speaker Affiliation
6.3.2012 Werner Kießling Augsburg University (WWW)
Abstract:

Preference SQL is a declarative extension of standard SQL by strict partial order preferences, developed at the University of Augsburg. In contrast to the exact-match query model of SQL, preferences queries behave like soft constraints under the Best-Matches-Only (BMO) query model. The Preference SQL query language extends standard SQL by introducing new language features for preferences using a unique intuitive and extensible preference constructor based approach. Qualitative preferences like e.g. Pareto / skyline as well quantitative preferences like numerical ranking, definable over categorical or numerical attribute domains can be used. Since Preference SQL is closed under strict-partial orders, all these constructors can be used inductively within one single complex Preference SQL query. To date, Preference SQL is the only approach supporting such as powerful comprehensive preference query model in database systems.

The current version of our Preference SQL System is implemented as a Java middleware component, enabling a seamless application integration with standard SQL back-end systems like e.g. Oracle, Postgres or MySQL. Our preference query optimizer performs algebraic transformations of preference relational algebra as well as cost-based algorithm selection for efficient Pareto / skyline evaluation. Currently we extend Preference SQL towards efficient support for mobile location-based spatial services and social networks.

This talk will cover selected aspect of Preference SQL, ranging from preference theory and modeling to preference query optimization issues and to context-aware preference generators for a commercial outdoor activity portal.

A trial version of the Preference SQL System can be accessed at www.trial.preferencesql.com.

Current trends in development and teaching of software engineering

Date Speaker Affiliation
1.11.2011 Vaclav Rajlich Department of Computer Science Wayne State University
Abstract:

This lecture reviews challenges and constrains that the instructor of a software engineering course faces. It argues that the best introduction into software engineering discipline is the training in the role of developers in a directed iterative process (DIP), where the most common task is software change (SC). In the course projects, students practice their skills by working on medium-sized open-source software systems, while the instructor supplies all the remaining DIP roles. A comprehensive overview of SC phases includes refactoring, concept location, impact analysis, unit testing, etc., and is the core of the course. At the end, the course briefly reviews the rest of the software engineering discipline. The results show that this organization of the course gives students a more realistic experience than traditional software engineering courses. The course has been taught repeatedly at Wayne State University and the students have expressed a high level of satisfaction. The resources required by such a course are comparable to other computer science courses. A new textbook supporting this approach is introduced [1].

[1] Vaclav Rajlich, Software Engineering: The Current Practice, CRC Press, 2011

Introduction to Similarity Search in Multimedia Databases [pdf]

Date Speaker Affiliation
19.04.2011 Tomas Skopal Department of Software Engineering, Charles University in Prague
Abstract:

In this lecture we give an introduction to the wide topic of similarity search in multimedia databases. Unlike table rows in a relational database, the objects of a multimedia database (i.e., images, audio, or video documents) have no explicitly defined structure of their content. Furthermore, unlike text documents or web pages that are composed of words, the multimedia content has no precise semantics. For example, a web page about car racings could be easily queried by a keyword query like "Ferrari AND winner". On the other hand, a collection of (not annotated) pictures taken on a car racing event cannot be searched so easily for "Ferrari" (or even "winner"), because the spots of red pixels possibly forming a Ferrari in the image is a knowledge that is understandable only to an experienced human. Therefore, searching for multimedia objects by its raw content is much more challenging task than searching a relational database, a text collection, or the Web. As a particular solution to this problem, the paradigm of content-based similarity search has been developed over the last two decades. We present the basic steps of the similarity search process: the feature extraction and similarity modeling, the querying, and finally the indexing for fast search.

Architecture in SOA [pdf (in Czech)]

Date Speaker Affiliation
05.04.2011 Michal Zemlicka Department of Software Engineering, Charles University in Prague
Abstract:

To appear...

Presentation of University of Luxembourg

Date Speaker Affiliation
30.03.2011 Franck Leprevost Computer Science and Communication Research Unit, University of Luxembourg
Abstract:

To appear...

Energy-Aware Scheduling for GreenIT in Large-Scale Distributed Systems [pdf1, pdf2]

Date Speaker Affiliation
30.03.2011 Pascal Bouvry Computer Science and Communication Research Unit, University of Luxembourg
Change of time: This seminar takes place on Wednesday, March 30 at 04:10PM in S8
Abstract:

There is a need to provide a holistic autonomic energy efficient solution to manage, provision, and administrate the various resources within large-scale distributed systems. We target Energy-aware Scheduling & Resource Allocations.We have investigated the energy efficiency and scheduling problem on scalable computing systems. We proposed an evolutionary algorithm based on a cGA + Local search and combined iterative random local search with DVS technique to minimize energy consumption without makespan degradation. The proposed solution outperforms related approaches in terms of energy and the completion time.

Bio:

Pascal Bouvry earned his undergraduate degree in Economical & Social Sciences and his Master degree in Computer Science with distinction (�91) from the University of Namur, Belgium . He went on to obtain his Ph.D. degree ('94) in Computer Science with great distinction at the University of Grenoble (INPG), France. His research at the IMAG laboratory focussed on Mapping and scheduling task graphs onto Distributed Memory Parallel Computers. Next, he performed post-doctoral research on coordination languages and multi-agent evolutionary computing at CWI in Amsterdam. Dr Bouvry gained industrial experience as manager of the technology consultant team for FICS (SONE) a world leader in electronic financial services. Next, he worked as CEO and CTO of SDC, a Saigon-based joint venture between SPT (a major telecom operator in Vietnam), Spacebel SA (a Belgian leader in Space, GIS and Healthcare), and IOIT, a public research and training center. After that, Dr Bouvry moved to Montreal as VP Production of Lat45 and Development Director for MetaSolv Software (ORCL), a world-leader in Operation Support Systems for the telecom industry (e.g. AT&T, Worldcom, Bell Canada, etc). Dr. Bouvry is currently heading the Computer Science and Communications (CSC) research unit of the Faculty of Sciences, Technology and Communications of Luxembourg University, and serving as Professor. Pascal Bouvry is also treasurer & member of the administration board of CRP-Tudor, and member of various scientific committees and technical workgroups (IEEE TCSC GreenIT steering committee, ERCIM WG, ANR, COST TIST, LIASIT, etc.)

Current Research Interests:
  • security, trust and reliability, for example, cryptology, intrusion detection, and reliable scheduling and routing on new generations of networks such as p2p, ad-hoc, and hybrids.
  • sustainable development, for instance, Energy Efficient Data Centers
  • systems biomedecine, for example, genomic sequencing, proteine folding, genomic modeling

Schema-less XML in Columns

Date Speaker Affiliation
22.03.2011 Jaroslav Pokorny Department of Software Engineering, Charles University in Prague
Abstract:

C-store environment uses a relational database for storing table tuples on the disk by columns. Can it be effectively used as XML database? The talk considers XML data without a schema. A two-level model of C-store based on XML-enabled relational databases is proposed. A measure of the model suitability is the possibility of evaluating effectively XPath queries. The XPath fragment considered allows the node-test not referring to attribute values and text values. Child, descendant, parent, ancestor, siblings, and following (preceding) are just the XPath axes used here. We will show that our algorithms are mostly of logarithmic complexity in n, where n is the number of nodes of XML tree associated with a XML document.

User Preference Ordering of Semantized Web Data [pdf]

Date Speaker Affiliation
08.03.2011 Peter Vojtas Department of Software Engineering, Charles University in Prague
Abstract:

Web Semantization is here understood as a process aimed to increase the amount of semantic data on the web. Main approach is automated (third party) annotation based on various web information extraction techniques. In this talk we discuss usefulness of semantized data for users. Our approach is based on R. Fagin's preference model. We mention some results in indexing and preference learning that enable Fagin model servers be used by many different users (interface, relational calculi, evaluation). We conclude with remarks on connections to software engineering lean start-up methodology and social networks.

Using Model Transformations to Describe Semantic Interoperability Mappings

Date Speaker Affiliation
14.06.2010 Jan Gottschick Fraunhofer-Institut fur Software- und Systemtechnik ISST Sichere Business IT-Infrastrukturen
Change of time: This seminar takes place on Monday, June 14 at 10:00AM in S1
Abstract:

The Web no longer defines itself through pre-build web pages simply delivered by a web server. Today, web-based services exchange data with Rich Internet Applications, which run completely inside the web browser or on mobile devices. The service front ends in the future will be gradually enriched by sophisticated mashups or apps. Further, the business processes between separate organisations have to be seamlessly integrated to exchange data without any media disruption. The service back ends are therefore designed to be highly reusable and feature a generic services pattern that requires well defined interfaces to provide a semantically interoperable data exchange.

Fraunhofer ISST developed the procedure model "Semantic Interoperability Engineering" (SIE) to improve the creation of high quality specifications defining a successful data exchange. This approach subsumes various actions to support the creation process, e.g. defining the basic concepts, developing strategies for solving semantic conflicts, defining meta specifications how to write compliant semantic interoperability specifications, defining a quality management system for the specifications, developing tools to automate the creation and compliance process, defining a technical publication process, and more.

This presentation gives a short overview about the semantic interoperability engineering and focusses on the "pivot mapping", which is the basic principle behind the overall approach. It motivates the idea to derive the required semantic mappings directly from platform independent and specific models based on the MDA methodology.

What is a Multi-Modeling Language?

Date Speaker Affiliation
21.04.2009 Martin Wirsing Ludwig-Maximilians-Universitaet Muenchen
Abstract:

In large software projects often multiple modeling languages are used in order to cover the different domains and views of the application and the language skills of the developers appropriately. Such "multi-modeling" raises many methodological and semantical questions, ranging from semantic consistency of the models written in different sublanguages to the correctness of model transformations between the sublanguages. We provide a first formal basis for answering such questions by proposing semantically well-founded notions of a multimodeling language and of semantic correctness for model transformations. In our approach, a multi-modeling language consists of a set of sublanguages and correct model transformations between some of the sublanguages. The abstract syntax of the sublanguages is given by MOF meta-models. The semantics of a multi-modeling language is given by associating an institution, i.e., an appropriate logic, to each of its sublanguages. The correctness of model transformations is defined by semantic connections between the institutions.

Computer-Assisted Proving for the Analysis of Systems and Specifications

Date Speaker Affiliation
17.03.2009 Wolfgang Schreiner Johannes Kepler University
Abstract:

Nowadays the formal methods landscape is dominated by model checking, an approach to the verification of programs/systems that is fully automatic but limited in its scope, in particular with respect to verifiable properties. In this talk, we discuss the alternative and more general direction of computer-assisted interactive proving which in the last decade (due to the development of automatic "Satisfiability Modulo Theories" solvers as supporting components) has made tremendous progress; consequently practical formal reasoning on more general system models and system properties has become viable. Furthermore, while not fully automatic and depending on human assistance, this approach has the potential of increasing a programmer's understanding (why is a system correct or what are the fundamental reasons of its failure) much more than the plain yes/no answers and counterexample executions produced by fully automatic model checkers. As an example of this direction, we present the "RISC ProofNavigator", an interactive proving assistant developed for education in reasoning about programs and specifications. Finally, we argue why modern computer science curricula should teach the use of proving assistants for formal specification and verification as an integral part of developing correct software.

Evaluating the Visual Syntax of UML:

Improving the Cognitive Effectiveness of the UML Suite of Diagrams

Date Speaker Affiliation
16.12.2008 Daniel L. Moody University of Twente
Abstract:

UML is a visual language. However surprisingly, there has been very little attention in either research or practice to the visual notations used in UML: both academic analyses and revisions to UML have focused almost exclusively on semantic issues, with little debate about or modification to the visual notations. We believe that this is a major oversight and that for this reason, UML's "visual development" is lagging behind its "semantic development". The lack of attention to visual aspects is particularly surprising given that the form of representations is known to have a comparatively greater effect on understanding and problem solving performance than their content. The UML visual notations were developed in a bottom-up manner, by reusing and synthesising existing notations. There is little or no justification for choice of graphical conventions, with decisions based on "expert consensus" rather than scientific evidence. This paper evaluates the UML family of diagrams using a set of predefined principles for visual notation design, which are based on theory and empirical evidence from communication, semiotics, graphic design, visual perception, psychophysics and cognitive science. This is the first comprehensive analysis of the UML visual language and covers all diagram types, constructs and symbols in the latest release of UML (2.1.2). The paper identifies some serious design flaws in the UML visual notations together with practical recommendations for improvement. The conclusion from the analysis is that radical surgery is required to the UML visual notations for them to be cognitively effective. We believe that the time is right for a major revamp of the UML visual notations now its semantics (the UML metamodel) is relatively complete and stable.

Software Architecture for Self-Managed Systems

Date Speaker Affiliation
11.11.2008 Jeff Magee Imperial College London
Abstract:

Self-management is put forward as one of the means by which we could provide systems that are scalable, support dynamic composition and rigorous analysis, and are flexible and robust in the presence of change. In the talk, we focus on architectural approaches to self- management, not because the language-level or network-level approaches are uninteresting or less promising, but because we believe that the architectural level provides the required level of abstraction and generality to deal with the challenges posed. A self-managed software architecture is one in which components automatically configure their interaction, in a way that is compatible with an overall architectural specification, to achieve the goals of the system. The objective is to minimise the degree of explicit management necessary for construction and subsequent evolution whilst preserving the architectural properties implied by its specification. The talk discusses some of the current promising work and presents an outline three-layer reference model as a context in which to articulate some of the main outstanding research challenges. In addition, the talk will describe a prototype system being developed at Imperial College that conforms to this model.

Software Architecture for Self-Managed Systems

Date Speaker Affiliation
17.10.2008 Hannu Jaakkola Tampere University of Technology Discussion topics in Software Process Improvement (SPI) - Research, Training and University level Education
Change of time: This seminar takes place on Friday, October 17th at 10:40AM in S4
Abstract:

The aim of the presentation is to introduce shortly the activities going on CoSE (Center of Software Expertise) Institute, which is a part of the Tampere University of Technology (Pori). The university itself is the second biggest technical university in Finland. It has a satellite faculty in the City of Pori as a part of University Consortium of Pori. CoSE Institute covers academic education, industry oriented training and research in the area of software systems. In research the focus of CoSE is in the problems of software development organisations - taking into account the quality and productivity. A special attention is given in the operations of small software organisations, which suffer from very much different improvement needs than the bigger ones. In addition, the well managed growth path of small software organisations (less than 10 employees) towards the category small / medium size (up to 50 employees) is not well known. There is also reasonable little amount of studies focusing in the problems of offshoring and well managed software processes in a distributed development organisation. The presentation introduces the work done by CoSE in the research area described above. The other academic activities of the organisation are introduced shortly, too. After the presentation discussion on co-operation activities between Charles University and Tampere University of Technology is possible.

An Introduction to Top-k and Skyline Computation

Date Speaker Affiliation
07.10.2008 Apostolos N. Papadopoulos Aristotle University of Thessaloniki
Change of place and time: This seminar takes place at 3:40PM in S3
Abstract:

Preference queries are very important to users, since they return the "best" objects according to some criteria. In this talk, we discuss the fundamental issues regarding Top-k and Skyline computation, which constitute the most important methods for determining the objects that best match users' preferences. More specifically, we study the fundamental research contributions by Ronald Fagin towards Top-k processing and an efficient Skyline algorithm. Moreover, we provide the necessary background material (e.g. R-trees) and give application examples where preference queries are of enormous interest. Additionally, we touch some advanced topics and briefly discuss research directions in the area.

Web 2.0 -- Core Concepts, Applications, Implications

Date Speaker Affiliation
07.10.2008 Tomas Pitner Masaryk University
Abstract:

Web 2.0 represents the recent evolutionary shift towards more user-oriented and user-driven web. The presentation will try to distil the essence of this concept - identify the distinguishing characteristics of the new web, illustrate them on successful services, and comment on the trends. As a specific topic, the phenomenon of integrated services - "mashups" - will be discussed, pointing out its crucial issues including technological as well as legal and business aspects.

Concept location

Date Speaker Affiliation
20.5.2008 Vaclav Rajlich Wayne State University
Abstract:

Concept location in source code is an evolution activity that identifies where a software system implements a specific concept. While it is well accepted that concept location is essential for the maintenance of complex procedural code like code written in C, it is much less obvious whether it is also needed for the maintenance of the Object-Oriented code. After all, the Object-Oriented code is structured into classes and well-designed classes already implement concepts, so the issue seems to be reduced to the selection of the appropriate class. The objective of our work is to see if the techniques for concept location are still needed (they are) and whether Object-Oriented structuring facilitates concept location (it does not).

Visions of Data Semantics: Another (and another) Look

Date Speaker Affiliation
12.05.08 Alex Borgida Rutgers University
Change of time: This seminar takes place on MONDAY May 12th, at 2PM (in S4 as usual)
Abstract:

The problem of data semantics is establishing and maintaining a correspondence between a data source (e.g., a database, an XML document) and its intended subject matter. We review the (relatively minor) role data semantics has played in Databases under the term "semantic data models", its more prominent place in ontology-based information integration, and then outline two new views: (i) Semantics as a composition of mappings between models, and (ii) Attaching intentional aspects (stakeholder goals) to Information Systems. (Joint work with John Mylopoulos and others at Univ. of Toronto)

(Organized jointly by DSE Colloquium and the project Intelligent Models, Algorithms, Methods and Tools for the Semantic Web Realisation)

MINIX 3: A Reliable and Secure Operating System

Date Speaker Affiliation
26.11.07 Andrew Tanenbaum VU University Amsterdam
Change of time and place: This seminar takes place on MONDAY November 26th, at 2PM in refectory 1st floor
Abstract:

Most computer users nowadays are nontechnical people and have a mental model of what they expect from a computer based on their experience with TV sets and stereos: you buy it, plug it in, and it works perfectly for the next 10 years. Unfortunately, they are often disappointed as computers are not very reliable when measured against the standards of other consumer electronics devices. A large part of the problem is the operating system, which is often millions of lines of kernel code, each of which can potentially bring the system down. The worst offenders are the device drivers, which have been shown to have bug rates 3-7x more than the rest of the system. As long as we maintain the current structure of the operating system as a huge single monolithic program full of foreign code and running in kernel mode, the situation will only get worse. While there have been ad hoc attempts to patch legacy systems, what is needed is a different approach. In an attempt to provide much higher reliability, we have created a new multiserver operating system with only 4000 lines in kernel and the rest of the operating system split up into small components each running as a separate user-mode process. For example, each device driver runs as a separate process and is rigidly controlled by the kernel to give it the absolute minimum amount of power to prevent bugs in it from damaging other system components. A reincarnation server periodically tests each user- mode component and automatically replaces failed or failing components on the fly, without bringing the system down and in some cases without affecting user processes. The talk will discuss the architecture of this system, called MINIX 3, The system can be downloaded for free from www.minix3.org.

Verifying Specifications with Proof Scores in CafeOBJ

Date Speaker Affiliation
13.11.07 Kokichi Futatsugi JAIST Japan Advanced Institute of Science and Technology
Abstract:

Verifying specifications is still one of the most important undeveloped topics in software engineering. It is important because quite a few critical bugs are caused at the level of domains,requirements, and/or designs. It is also important for the cases where no program codes are generated and specifications are analyzed/ verified only for justifying models of problems in real world. In this talk, a survey of our research activities in verifying specifications is given. After explaining fundamental issues and importance of verifying specifications, the proof score approach in CafeOBJ and its applications to several areas are described.

Modeling and Architecture of an Integrative Environmental Simulation System

Date Speaker Affiliation
30.10.07 Rolf Hennicker LMU Muenchen
Abstract:

We describe the concepts and design principles of an environmental simulation system which supports the study and analysis of water-related global change scenarios in the Upper Danube Basin. The system provides a Web-based platform integrating the distributed simulation models of all socio-economic and natural science disciplines taking part in the GLOWA-Danube project which is part of the German programme on global change in the hydrological cycle. Crucial aspects of the system development concern the specification of interfaces between simulation models, the treatment of the simulation space, the modeling of socio-economic actors and the coordination of coupled simulations for which we have developed a coordination framework. To ensure the correctness of the synchronization of concurrently running simulation models we have applied formal methods of process algebra.

Enterprise Content Management

Date Speaker Affiliation
16.10.07 Peter Eklund The University of Wollongong
Abstract:

We deal with Enterprise Content Management and some of the tough computing issues that are faced in that domain. This mostly draws on my experience working with my secondment company, www.objective.com, and is a mixed bag of problems: database replication, information retrieval, object caching, security and access control.

Automatic Feedback-Directed Object Inlining in the Java HotSpot(TM) Virtual Machine

Date Speaker Affiliation
19.06.07 Hanspeter Moessenboeck Johannes Kepler University Linz
Change of place: This seminar takes place in the room S1
Abstract:

Object inlining is an optimization that embeds certain referenced objects into their referencing objects. It reduces the access costs of those objects by eliminating unnecessary loads of references. We change the order of objects in the heap in such a way that objects that are frequently accessed together are placed next to each other in memory so that their distance is fixed, i.e. these objects are colocated. This allows field loads to be replaced by address arithmetic. We implemented this optimization for Sun Microsystems' Java HotSpot VM. The necessary analysis is performed automatically at run time, requires no actions on the part of the programmer and supports dynamic class loading. We use read barriers to detect the most frequent references between objects, i.e. the ones that are worth being optimized. If an object A has a field f referencing an object B the loading of f can be safely eliminated if A and B are colocated and if the following two preconditions are satisfied: A and B must be allocated together, and f must not be changed later. These preconditions are checked by the just-in-time compiler avoiding a global data flow analysis. The garbage collector ensures that groups of colocated objects are not split: it copies groups as a whole to their new locations. The evaluation shows that our dynamic approach successfully identifies and optimizes frequently accessed fields for several benchmarks. The average peak performance speedup of 9% for SPECjvm98 (with a maximum speedup of 51%) justifies the startup overhead of 3% on average (with a maximum slowdown of 11%) that is caused mainly by the read barriers and the additional compilation of methods.

Teaching Evolution of Open-Source Projects in Software Engineering Courses

Date Speaker Affiliation
29.5.2007 Vaclav Rajlich Wayne State University
Change of place: This seminar takes place in the room S1
Abstract:

In the traditional software engineering courses, the students develop small programs from scratch. This does not correspond to industry practice where programmers spend most of their time evolving medium to large systems. In order to narrow this gap, we developed a course where students practice software evolution through the implementation of change requests on medium-sized open-source software systems. The results of the course show that this type of software engineering course gives students a more realistic experience than traditional software engineering courses. In the survey at the end of the course, the students expressed a higher level of satisfaction with both rating the course and assessing how much they learned. Additionally, the resources required by such a course are not excessive.

On Teaching

Date Speaker Affiliation
15.5.2007 Petr Tuma DSE
Abstract:

The seminar will present a collection of various general issues related to teaching, mostly along the lines of what is the purpose of teaching, what are the common misconceptions of teaching among teachers, what is useful and what is useless to do when trying to improve teaching. The sources used include work by Paul Ramsden on university teaching and work by Joseph Novak on understanding knowledge.

Composite Contracts

Date Speaker Affiliation
24.4.2007 Nicolas Rivierre France TÚlÚcom
Abstract:

Abadi and Lamport established a general theorem for composing specifications. Based on an assume-guarantee principle, it enables one to prove the specification of a composite system from the ones of its components. In this talk, I'll investigate how this result can be exploited to leverage an existing contracting framework for hierarchical software components, in which contracts are first-class objects during configuration and run times. This framework already associates specifications and responsibilities to software components, within the traditional horizontal form of composition. We will show here how the vertical one can be made operational using the theorem as a sound formal basis.

The content of this web site is licensed under Creative Commons Attribution-NonCommercial 3.0 Czech Republic