Hauptnavigation

News from the Artificial Intelligence Group

The chair of artificial intelligence deals with the wide field of machine learning. In particular the chair concentrates on the development and implementation of learning algorithms that solve challenging problems.

Ausschreibung zur Vertretungsstelle im Sekretariatsbereich

An der TU Dortmund, Fakultät für Informatik am Lehrstuhl VIII, ist zum nächstmöglichen Zeitpunkt die Stelle einer Sachbearbeiterin/eines Sachbearbeiters im Sekretariatsbereich befristet zur Vertretung zu besetzen. (more...  )

Celebratory Colloquium at the Faculty for Computer Science

Katharina Morik

Ein besonderes Spektrum an Vorträgen fand am Jahresende zum 60. Geburtstag von Katharina Morik statt. Gemeinsam war den drei Hauptrednern, dass sie im Bereich Maschinelles Lernen bzw. Data Mining international höchst renommiert sind und bei Katharina Morik an der (Technischen) Universität Dortmund promovierten. Völlig verschieden ihre Tätigkeitsfelder.

Inhaltliche Gemeinsamkeiten der Redner mit der Jubilarin wurden in der kurzen Einführung deutlich, in der Katharina Morik ihre Forschungsziele zusammenfasste, die sie an der TU Dortmund verfolgt: situierte Systeme, die durch Lernfähigkeit Sensorik, Kommunikation und Handlung verbinden. Anfang der 90er Jahre entstanden Arbeiten zur Robotik: realzeitlich wurden in verteilten, heterogenen Datenströmen Muster entdeckt, die zur Handlungsplanung eingesetzt wurden. Der SFB 876 (Informatik), dessen zweite Phase gerade bewilligt wurde, kann mit seiner Verbindung von Datenanalyse und Cyber Physical Systems in der Leitlinie lernfähiger, situierter Systeme gesehen werden.

Die Arbeiten zu sehr großen Datenmengen, die Katharina Morik in 12 Jahren im SFB 475 (Statistik) zusammen mit Claus Weihs durchgeführt hat, wurden von der Sprecherin dieses Sonderforschungsbereichs, Ursula Gather, in einer kurzen Ansprache gewürdigt.

Ganz unterschiedliche Herangehensweisen, maschinelles Lernen erfolgreich zu erforschen und anzuwenden, wurden durch die Hauptvorträge der drei herausragenden Wissenschaftler deutlich.

Stefan Wrobel Thorsten Joachims Ingo Mierswa

Unsere Studierenden mag es freuen, wenn sie einige Beispiele sehen, wozu das Studium an der TU Dortmund befähigt: Forschungsdirektor, Professor, CEO einer Firma – auf der Grundlage herausragender Forschung zu maschinellem Lernen, Data Mining, Big Data Analytics lässt sich einiges machen!

 

(more...  )

Merry Christmas and a happy new year

Christmas 2014

From 19 of December 2014 until 9th of January 2015 the secretary's office is not occupied.
We wish you a merry christmas and a happy new year!

 

LS8 publishes SpringerBrief on boosting statistical relational learners

This SpringerBrief addresses the challenges of analyzing multi-relational and noisy data by proposing several Statistical Relational Learning (SRL) methods. These methods combine the expressiveness of first-order logic and the ability of probability theory to handle uncertainty. It provides an overview of the methods and the key assumptions that allow for adaptation to different models and real world applications. The models are highly attractive due to their compactness and comprehensibility but learning their structure is computationally intensive. To combat this problem, the authors review the use of functional gradients for boosting the structure and the parameters of statistical relational models. The algorithms have been applied successfully in several SRL settings and have been adapted to several real problems from Information extraction in text to medical problems. Including both context and well-tested applications, Boosting Statistical Relational Learning from Benchmarks to Data-Driven Medicine is designed for researchers and professionals in machine learning and data mining. Computer engineers or students interested in statistics, data management, or health informatics will also find this brief a valuable resource.

(more...  )

Stellen für studentische Hilfskräfte ab Januar 2015

An der TU Dortmund, Fakultät für Informatik am Lehrstuhl VIII, sind ab Januar 2015 Stellen für Studentische Hilfskräfte zu besetzen.

(more...  )

Vorlesung Natürlichsprachliche Systeme (Katharina Morik)

IBM Watson

Google, Facebook oder Netflix brauchen für viele ihrer Dienste die Verarbeitung natürlicher Sprache. So gibt es die große Abteilung Natural Language Processing bei Google http://research.google.com/pubs/NaturalLanguageProcessing.html

Das IBM-Programm Watson konnte im Februar 2011 in dem Quiz Jeopardy auf natürlichsprachliche Fragen besser antworten als zwei menschliche Quiz-Sieger.

Ray Kurzweil (Google Director of Engineering) möchte darüber hinausgehen: „So IBM’s Watson is a pretty weak reader on each page, but it read the 200m pages of Wikipedia. And basically what I'm doing at Google is to try to go beyond what Watson could do.“ http://searchengineland.com/ray-kurzweils-job-google-beat-ibms-watson-natural-language-search-185149 Es gibt eine Fülle von Methoden zur Analyse sehr großer Textmengen für ebenfalls viele Anwendungen: Sentiment Analysis, personalisierte Werbung, Empfehlungen, email Routing, automatische Texterstellung für Kurznachrichten und Reporting, automatische Fragebeantwortung, Informationsextraktion aus dem WWW. In der Vorlesung mit Übungen lernen Sie die Methoden und Werkzeuge dazu kennen. Das neue Lehrkonzept beinhaltet inverted class room Sitzungen und selbstständige Arbeiten, so dass Sie für die Praxis gerüstet sind. http://www-ai.cs.uni-dortmund.de/LEHRE/VORLESUNGEN/NLS/WS1415/index.html

Vorlesung Probabilistische Graphische Modelle (Kristian Kersting)

Wie handelt man unter Unsicherheit, bei fehlenden oder fehlerhaften Daten? Um mit solchen Unsicherheiten umgehen zu können, haben sich in den letzten Jahren probabilistische, graphischen Modellen bewährt. Sie gehören zu den Bemühungen der modernen Informationstechnik, das Schlussfolgern unter Unsicherheit zu ermöglichen.

Tag-Cloud Probabilistische graphische Modelle

Prominente Anwendungsfelder sind die Robotik, die Bioinformatik, die Künstliche Intelligenz, das Maschinelle Lernen. So kommen sie zum Beispiel in der Auswertung von medizinischen Daten, der Analyse von Genexpressionsdaten und dem Tracken von Bewegungen zum Einsatz. Gegenstand der Vorlesung "Probabilistische Graphische Modelle" des LS8 sind grundlegende Fragestellungen und Techniken der graphischen Modelle. http://www-ai.cs.uni-dortmund.de/LEHRE/VORLESUNGEN/PGM/WS1415/index.html

Vorlesung Large-Scale Optimization (Sangkyun Lee)

Optimierung

Ganz allgemein sind Daten oft billiger zu erhalten als das Wissen von Experten zu extrahieren und dann zu modellieren. Aber wie können Rechner automatisch große Modelle --- wie sie in der Verarbeitung natürlicher Sprache, bei dem Schätzen von Graphischen Modellen und im statischen Maschinellen Lernen auftreten --- aus Daten schätzen?

In den meisten Lernverfahren steckt als Kern eine Optimierungsaufgabe: der Fehler soll miniert oder die Wahrscheinlichkeit für das richtige Ergebnis maximiert werden. Die theoretischen Grundlagen und Methoden behandelt in englischer Sprache die Vorlesung "Large-Scale Optimization".

PG infoscreen (Kristian Kersting, Hendrik Blom)

Infoscreen

Die Ansätze aus allen Vorlesungen können dann zur Anwendungen in der PG "Infoscreen" kommen. Infoscreens sind digitale Bildflächen und sollen eine besondere Aufmerksamkeit in "reizarmen" öffentlichen Räumen erzielen.

Es soll über Aktuelles an der Fakultät für Informatik der TU Dortmund informiert werden.

KDD 2014 sold out

KDD 2014 is sold out. They had to close registrations. 2200 attendees will enjoy the conference next week in Times Square. Katharina Morik gives a keynote talk at the workshop BigMine’14.

(more...  )

The virtual steel works

After the press conference the LS8 project (Katharina Morik, Hendrik Blom, Tobias Beckers)  in collaboration with the SMS Siemag and the Dillinger Hütte is outlined in two interviews: Dominik Schöne of the Dillinger Hütte and Katharina Morik.

(more...  )

ViSTA-TV in a Nutshell

The European project VistaTV had its successful final review meeting in Amsterdam, 1st of July. LS 8 contributed live stream analysis separating ads from shows in internet television. Online recommendations of shows based on user behavior have been produced based on Termset Clustering.

 

(more...  )

Mediaday of the SMS Group in Hilchenbach

Mediaday of the SMS Group in Hilchenbach at 3. July 2014

Data Mining/ Industrie 4.0

Summary talk by Katharina Morik about "Data Mining, Big Data and Prediction Models"

(more...  )

Talk at the TU Dortmund: What happens to our data? Between permanent harassment paranoia and post-privacy

Wednesday, 2. July 2014, 16:00 (s.t.) -18:30, P1-05-309
  • Kristian Kersting (Chair for artificial intelligence)
  • Sarah Küsgen (Chair for service and technology management)
  • Kai-Uwe Loser (Data security engineer of the RUB)
  • Johannes Weyer & Robin D. Fink (specific field technical sociology)


The youngest exposures of whistle-blowser Edward Snowden showed one more time the attractiveness of collecting massive data in the age of social media.

The question 'what happens to our data?', viewed from technical, economic and sociological background, will be investigated in the context of this event. The technical possibilities of modern data-mining are diverse and allow conclusions down to the individual level. Collected data from social networks are especially attractive for marketing and product design. Behind this background the protection of privacy will be assigned to new tasks.

The contributors will hold a 10-15 minutes talk each and will afterwards take part in a discussion with the audience. The event will be moderated by Johannes Weyer.

(more...  )

Moving to Otto-Hahn-Str. 12

The chair for Artificial Intelligence is moving to the new building in Otto-Hahn-Str. 12. Thus, between 06/30/14 and 07/04/14 we may not be available at all times.

Talk at VigLink: Resource-aware graphical models

Prof. Morik talks at VigLink

Abstract:
Machine learning can help to enhance small devices. For instance, keeping the energy consumption of smart phones low is one of the major concerns of the users, as is well illustrated by various “charge your mobile” stations at public places. Where the operating systems of smart phones already offer heuristics and battery apps show consumption profiles, machine learning can do more. Predictions allow better optimizations of the operating system, prepare for particular app usages at certain points in time, or manage services such as GPS or WLAN in a context-aware and adaptive manner. This challenges learning algorithms to real-time application of their models. Moreover, it demands the models to run on the resource-restricted device without consuming more energy themselves than they save!

(more...  )

Talk at NASA: Data Analytics for Sustainability

Title: Data Analytics for Sustainability

  • Speaker: Katharina Morik, Technische Universität, Dortmund
  • Date & Time: Wednesday, May 28, 2:00 pm - 3:00 pm
  • Location: Building N245 Auditorium

Abstract:

Sustainability has many facets and researchers from many disciplines are working onthem. Particularly knowledge discovery always considered sustainability an importanttopic (e.g., special issue on data mining for sustainability in Data Mining andKnowledge Discovery Journal, March 2012).

Host: Dr. Kamalika Das
NASA Ames Research Center
MS 269-1, PO Box 1, Moffett Field, CA 94035

PROF. MORIK continues her series of lectures at google

 

On Tue 05/27/2014 Prof. Katharina Morik give a talk about "Resource-aware graphical models and spatio-temporal predictions" at the Google Headquarters in Palo Alto, California, USA.

Abstract:
Machine learning can help to enhance small devices. For instance, keeping the energy consumption of smart phones low is one of the major concerns of the users, as is well illustrated by various “charge your mobile” stations at public places. Where the operating systems of smart phones already offer heuristics and battery apps show consumption profiles, machine learning can do more. Predictions allow better optimizations of the operating system, prepare for particular app usages at certain points in time, or manage services such as GPS or WLAN in a context-aware and adaptive manner. This challenges learning algorithms to real-time application of their models. Moreover, it demands the models to run on the resource-restricted device without consuming more energy themselves than they save!
In the talk, graphical models are presented that face these challenges. Using Conditional Random Fields (CRF) for the prediction of files that the user will fetch next on her smart phone can be used by the operating system for organizing the memory. Analyzing groups of apps running on the smart phone may estimate the energy consumption over time.
A novel spatio-temporal random field (STRF) has been implemented, smoothing the temporal changes and distributing the optimization. This graphical model has been used to predict app usage over time. In another application, it has been combined with a trip planner resulting in smart routing for smart cities. In order to run graphical models on very restricted devices, even those withoutvfloating point calculation, one computing with integer values only has been developed. The integer approximation of graphical models shows good accuracy and speed-up and opens up novel applications on resource-restricted devices.

PROF. MORIK gave A TALK ABOUT 'DATA ANALYTICS FOR SUSTAINABILITY' AT THE Cornell University in New York, USA

 

Sustainability has many facets and researchers from many disciplines are working on them. Particularly knowledge discovery always considered sustainability an important topic (e.g., special issue on data mining for sustainability in Data Mining and Knowledge Discovery Journal, March 2012).

(more...  )

Prof. Morik gives a talk about 'Data Analytics for Sustainability' at the University of Maryland, Baltimore County on Thursday 22 May 2014.

 

Sustainability has many facets and researchers from many disciplines are working on them. Particularly knowledge discovery always considered sustainability an important topic (e.g., special issue on data mining for sustainability in Data Mining and Knowledge Discovery Journal, March 2012).

  • Environmental tasks include risk analysis concerning floods, earthquakes, fires, and other disasters as well as the ability to react to them in order to guarantee resilience. The climate is certainly of influence and the debate on climate change received quite some attention.
  • Energy efficiency demands energy-aware algorithms, operating systems, green computing. System operations are to be adapted to a predicted user behavior such that the required processing is optimized with respect to minimal energy consumption.
  • Engineering tasks in manufacturing, assembly, material processing, and waste removal or recycling offer opportunities to save resources to a large degree. Adding the prediction precision of learning algorithms to the general knowledge of the engineers allows for surprisingly large savings.

Global reports on the millennium goals and open government data regarding sustainability are publicly available. For the investigation of influence factors, however, data analytics is necessary. Big data challenges the analysis to create data summaries. Moreover, the prediction of states is necessary in order to plan accordingly. In this talk, two case studies will be presented. Disaster management in case of a flood combines diverse sensor data streams for a better traffic administration. A novel spatiotemporal random field approach is used for smart routing based on traffic predictions. The other case study is in engineering and saves energy in the steel production based on the multivariate prediction of the processing end-point by the regression support vector machine.

11:00am-12:30pm, Thursday 22 May 2014, ITE 456, UMBC

(more...  )

Call for Papers - MLDM 2015

MLDM 2015

11th International Conference on Machine Learning and Data Mining

July 11 - 24, 2015, Freie Hansestadt Hamburg, Germany

This congress will feature three events the 11th International Conference on Machine Learning and Data Mining MLDM, the 15 th Industrial Conference on Data Mining ICDM ( www.data-mining-forum.de), and the 10 th International Conference on Mass Data Analyisis of Signals and Images MDA (www.mda-signals.de). Workshops and Tutorial will also be given.

  • Submission of papers: January 15th, 2015
  • Notification of acceptance: February 28, 2015
  • Submission of camera-ready copy: April 5th, 2015
(more...  )

Katharina Morik at the University Vienna

Dortmunder postdoc Wouter Duivesteijn wins C.J. Kok Jury Award 2013.
Prof. Dr. Dr. h. c. Monika Henzinger und Prof. Dr. Katharina Morik with some participants of the college, where Katharina Morik gives a course “Data Analytics”.

More than 1 year after the faculty of computer science at the TU Dortmund has conferred an honorary doctorate to Monika Henzinger, Professor at the University of Vienna, Katharina Morik gives a course on "Data Analytics" in the context of the interdisciplinary college at the computer science of the University of Vienna and also presented in a well-attended colloquium lecture results of the SFB876: "Big Data Analytics and Astrophysics".

Workshop: Needles In a Stream of Hay (NISH2014)

Workshop collocated with INFORMATIK 2014, September 22-26, Stuttgart, Germany.

This workshop focuses on the area where two branches of data analysis research meet: data stream mining, and local exceptionality detection.

Local exceptionality detection is an umbrella term describing data analysis methods that strive to find the needle in a hay stack: outliers, frequent patterns, subgroups, etcetera. The common ground is that a subset of the data is sought where something exceptional is going on: finding the needles in a hay stack.

Data stream mining can be seen as a facet of Big Data analysis. Streaming data is not necessarily big in terms of volume per se but instead it can be in terms of the high troughput rate. Gathering data for analyzing is infeasible so the relevant data of a data point has to be extracted when it arrives.

Submission

Submissions are possible as either a full paper or extended abstract. Full papers should present original studies that combine aspects of both the following branches of data analysis:

stream mining: extracting the relevant information from data that arrives at such a high throughput rate, that analysis or even recording of records in the data is prohibited;
local exceptionality mining: finding subsets of the data where something exceptional is going on.

In addition, extended abstracts may present position statements or results of original studies concerning only one of the aforementioned branches.

Full papers can consist of a maximum of 12 pages; extended abstracts of up to 4 pages, following the LNI formatting guidelines. The only accepted format for submitted papers is PDF. Each paper submission will be reviewed by at least two members of the program committee.

(more...  )

NEM Position Paper of Big and Open Data

"NEM position papers are documents giving the NEM Initiative view on any subject related to the networked electronic media area. The NEM position papers typically include: letters of advice to the Commission, formal opinions submitted to the Commissioner, submissions to regulatory bodies, or any other formal statement of this nature, as well as further views of the NEM community on various technological, societal, and policy issues related to NEM." Source: www.nem-initiative.org

(more...  )

Many companies hope for big data

Our students at LS 8 learn exactly what is in demand at many companies.

(more...  )

Dortmunder postdoc Wouter Duivesteijn wins C.J. Kok Jury Award 2013.

Dortmunder postdoc Wouter Duivesteijn wins C.J. Kok Jury Award 2013.

Annually, the Faculty of Science at Leiden University, the Netherlands, grants the C.J. Kok Jury Award for the best PhD thesis of the past year. All institutes within the faculty (astronomy, physics, mathematics, computer science, chemistry, pharmacy, biology, and environmental sciences) are given the opportunity to nominate candidates for the award.

 Out of a pool of over 120 dissertations, the C.J. Kok Jury Award 2013 was won by Wouter Duivesteijn, with his thesis "Exceptional Model Mining". Notably, this is the first time ever that the award (existing since 1971) has been bestowed upon a computer scientist.

Show news archive