Fachgruppe

Knowledge Discovery, Data Mining and Machine Learning

Programm Fachgruppe KDML

Day 1

13:00 – 13:30 Uhr
Opening
Prof. Dr. Stefan Wrobel (Institute Director Fraunhofer IAIS)

13:30 – 15:00 Uhr
Keynote
Prof. Dr. Geoff Webb (Monash University)

Time series classification at scale

Time series classification is a fundamental data science problem, providing understanding of dynamic processes as they evolve over time. The recent introduction of ensemble techniques has revolutionised this field, greatly increasing accuracy, but at a cost of increasing already burdensome computational overheads. I present new time series classification technologies that achieve the same accuracy as recent state-of-the-art developments, but with many orders of magnitude greater efficiency and scalability. These make time series classification feasible at hitherto unattainable scale.

15:00 – 15:30 Uhr
Break

15:30 – 16:30 Uhr
Joint Research Track

Solving Abstract Reasoning Tasks with Grammatical Evolution
(Raphael Fischer, Matthias Jakobs, Sascha Mücke and Katharina Morik)

Construction of a Corpus for the Evaluation of Textual Case-based Reasoning Architectures
(Andreas Korger and Joachim Baumeister)

Modeling Interdependent Preferences over Incomplete Knowledge Graph Query Answers
(Till Affeldt, Stephan Mennicke and Wolf-Tilo Balke)

16:30 – 18:00 Uhr
Poster Session

Day 2

13:00 – 14:30 Uhr
Keynote
Prof. Dr. Kristian Kersting (TU Darmstadt)

On Hybrid and Systems AI
Our minds make inferences that appear to go far beyond standard machine learning. Whereas people can learn richer representations and use them for a wider range of learning tasks, machine learning algorithms have been mainly employed in a stand-alone context, constructing a single function from a table of training examples. In this talk, I shall touch upon a view on AI and machine learning, called Systems AI, that can help capturing these human learning aspects by combining different AI and ML models using high-level programming. Since inference remains intractable, existing approaches leverage deep learning for inference. Instead of “just going down the neural road,” I shall argue to also use probabilistic circuits, a deep but tractable architecture for probability distributions. This hybrid approach can speed up inference as I shall illustrate for unsupervised science understanding, database queries and automating density estimation.

14:30 – 15:30 Uhr
Clustering and Time Series

Grace – Limiting the Number of Grid Cells for Clustering High-Dimensional Data
(Anna Beer, Daniyal Kazempour, Julian Busch, Alexander Tekles and Thomas Seidl)
Segmenting and Clustering Noisy Arguments
(Lorik Dumani, Christin Katharina Kreutz, Manuel Biertz, Alex Witry and Ralf Schenkel)
(SIAM SDM 2020) Two-Sample Testing for Event Impacts in Time Series
(Erik Scharwächter and Emmanuel Müller)

15:30 – 16:00 Uhr
Break

16:00 – 17:00
Neural Networks

Quality Guarantees for Autoencoders via Unsupervised Adversarial Attacks
(Benedikt Böing, Roy Rajarshi, Emmanuel Müller and Daniel Neider)
Combining Universal Adversarial Perturbations
(Beat Tödtli and Maurus Kühne)
Phantom Embeddings: Using Embeddings Space for Model Regularization in Deep Neural Networks
(Mofassir Ul Islam Arif, Mohsan Jameel, Josif Grabocka and Lars Schmidt-Thieme)
The Shape of Data: Intrinsic Distance for Data Distributions
(Anton Tsitsulin, Marina Munkhoeva, Davide Mottin, Panagiotis Karras, Alexander Bronstein, Ivan Oseledets and Emmanuel Müller)

17:00 – 18:00 Uhr
Community Meeting

Day 3

13:00 – 14:00 Uhr
Applications

A hierarchical multi-level product classification workbench for retail
(Maximilian Harth, Christian Schorr and Rolf Krieger)
Why did my Consumer Shop? Learning an Efficient Distance Metric for Retailer Transaction Data
(Yorick Spenrath, Marwan Hassani, Boudewijn Van Dongen and Haseeb Tariq)
Comparison of knowledge based feature vector extraction and geometrical parameters of Photovoltaic I-V Curves
(Cem Basoglu, Grit Behrens, Konrad Mertes and Matthias Diehl)
Using Probabilistic Soft Logic to Improve Information Extraction in the Legal Domain
(Birgit Kirsch, Sven Giesselbach, Timothée Schmude and Stefan Rüping)

14:00 – 15:00
Emotions and Semantics

Fusing Multi-label Classification and Semantic Tagging
(Jörg Kindermann and Katharina Beckh)
Native sentiment analysis tools vs. translation services – Comparing GerVADER and VADER
(Karsten Tymann, Louis Steinkamp, Oxana Zhurakovskaya and Carsten Gips)
EmoDex – An emotion detection tool composed of established techniques
(Oxana Zhurakovskaya, Louis Steinkamp, Karsten Tymann and Carsten Gips)

15:00 – 15:30 Uhr
Break

15:30 – 16:30 Uhr
Keynote
Prof. Dr. Thomas Gärtner (TU Wien)

Interactive Machine Learning with Structured Data

In this talk I’ll give an overview of our contributions to what I call interactive machine learning. Often, interaction in Computer Science is interpreted as the interaction of humans with the computer but I intend a broader meaning of the interaction of machine learning algorithms with the real world, including but not restricted to humans. Interactions with humans span a broad range where they can be intentional and guided by the human or they can be guided by the computer such that the human is oblivious of the fact that he is being guided. Another example of an interaction with the real world is the use of machine learning algorithms in cyclic discovery processes such as drug design. Important properties of interactive machine learning algorithms include efficiency, effectiveness, responsiveness, and robustness. In the talk I will show how these can be achieved in a variety of interactive contexts.

16:30 – 17:00 Uhr
Closing

Call for Papers

→ Fachgruppe Knowledge Discovery, Data Mining and Machine Learning

Paper Submission: We are accepting late breaking submissions until July 14, 2020

KDML is a workshop series that aims at bringing together the German Machine Learning and Data Mining community. The KDML 2020 Workshop is co-located with the annual LWDA 2020 – Learning, Knowledge, Data, and Analysis – conference and will take place from September 09, 2020 to September 11, 2020, at the Rheinische Friedrich-Wilhelms-Universität Bonn.

We invite submissions on all aspects of data mining, knowledge discovery, and machine learning. In addition to original research, we also invite resubmissions of recently published articles at major conference venues related to KDML. Moreover, KDML explicitly invites student submissions. Topics of interest cover foundations and applications of all areas of data mining and machine learning. If you would like to submit a paper on a topic and are in doubt about its relevance, please contact the workshop organizers.

Topics of interest

Topics of interest include but are not limited to:

Foundations, models, and theory of machine learning and data mining
Supervised, semi-supervised, and unsupervised learning
Rule-based learning and pattern mining
Multi-objective learning
Deep learning
Safety related aspects in Deep Learning
Explainability in neural networks
Representation and embedding learning
Time series; spatiotemporal data mining, mining sequences, stream mining
Unstructured, semi-structured, multi-modal data mining
Network, graph, and Web mining
Parallel and Distributed data mining
Applications of data mining and machine learning in all domains including healthcare, financial sector, environment, engineering, the Web
Open source frameworks and tools for data mining and machine learning

Types of submissions

We solicit new contributions (up to 12 pages, peer-reviewed and to be published by LWDA). Shorter contributions (4 pages) are also solicited. We welcome submissions in English and German, however English is preferred. All papers must be formatted according to the Springer LNCS guidelines. All contributions must be submitted via EasyChair using the link: https://easychair.org/conferences/?conf=lwda2020 ; only PDF is permitted. Please select the track „FG-KDML’“ for your submission. We further welcome submissions of works accepted recently at top-tier international venues related to KDML (e.g., KDD, ECML, ICML, NIPS, IJCAI, AAAI, ICDM, SDM, etc.). These will not be reviewed but selected by the PC chairs. They will not be included in the LWDA proceedings.

Evaluation process and publication of unpublished submissions

Each submission will be reviewed by at least two independent reviewers.
The conference proceedings will be published as CEUR Workshop Proceedings and will be indexed by DBLP.

Participation and presentation

All workshop participants have to register for the LWDA 2020 conference. Papers will be accepted for either long (30 min) or short (10 min) presentations; authors are expected to also prepare a poster presentation of their work.

Important dates

Late breaking submission deadline: July 14, 2020
Notification of acceptance: July 27, 2020
Camera-ready copy: August 17, 2020
LWDA 2020 Conference: September 09 – September 11, 2020

Intended audience

The target group includes researchers and practitioners who are interested in developing, applying and analyzing knowledge and experience management systems as well as applicable scenarios. The workshop is also a great and affordable platform for young researchers to present their work to a larger group of researchers and get valuable feedback.

Workshop chair

Dr. Pascal Welke, Rheinische Friedrich-Wilhelms-Universität Bonn
Dr. Nico Piatkowski, Fraunhofer-Institut für Intelligente Analyse- und Informationssysteme IAIS

Program committee

Maram Akila, Fraunhofer
Klaus-Dieter Althoff, DFKI / University of Hildesheim
Martin Atzmueller, Tilburg University
Christian Bauckhage, Fraunhofer
Rainer Gemulla, University of Mannheim
Stephan Günnemann, Technical University of Munich
Sibylle Hess, Data Mining Group, TU Eindhoven
Andreas Hotho, University of Wuerzburg
Sebastian Houben, Fraunhofer
Marius Kloft, TU Kaiserslautern
Christian Kühnert, Fraunhofer
Florian Lemmerich, RWTH Aachen University
Thomas Liebig, Materna SE
Michael Mock, Fraunhofer
Petar Ristoski, IBM Research-Almaden
Ute Schmid, University of Bamberg
Thomas Seidl, Ludwig-Maximilians-University (LMU) Munich
Stefan Wrobel, Fraunhofer & University of Bonn

Chair

Dr. Daniel Trabold

Fraunhofer-Institut für Intelligente Analyse- und Informationssysteme IAIS

Schloss Birlinghoven
53757 Sankt Augustin, Germany

Phone +49 2241 14-2751