Important Dates:
Application opens:	March 1, 2018
Application closes:	May 15, 2018
Notification of acceptance:	May 20, 2018
Deadline for payment of registration:	June 1, 2018
Arrival of participants:	July 1, 2018
Start of event:	July 2, 2018
End of event:	July 6, 2018

Eighth European Business Intelligence & Big Data Summer School (eBISS 2018)

Invited Speakers & Tutors

Boudewijn van Dongen

Eindhoven University of Technology, The Netherlands

Boudewijn’s research focusses on conformance checking. Conformance checking is considered to be anything where observed behavior, needs to be related to already modeled behavior. Conformance checking is embedded in the larger contexts of Business Process Management and Process Mining. Boudewijn aims to develop techniques and tools to analyze databases and logs of large-scale information systems for the purpose of detecting, isolating, diagnosing and predicting misconformance in the business processes supported by these systems. The notion of alignments play a seminal role in conformance checking and the AIS group is world-leading in the definition of alignments for various types of observed behavior and for various modelling languages.

Email: B.F.v.Dongen@tue.nl
Web:

Lecture: Process Mining: Are we doing things right?
The challenge in process mining is to obtain insights into processes through the structured and automatic analysis of event data. Event data shows who performed what action in which context at what point in time. Event data appears in many forms and types and various techniques exist to obtain models from this data.
However, in many cases, models are also made by professionals. For example, in software engineering, architectural models are made by software architects. Another example are the models made for ISO 9001 certification of business processes by consultants.
Conformance checking is considered to be anything where observed behavior needs to be related to already modeled behavior. In other words, if models are given and event data is recorded, how can they be related and how can we obtain insights into deviations?
In this lecture, I discuss the basics of conformance checking using models and event logs as input. I also discuss the state-of-the-art in tool support.

Anne Rozinat

Fluxicon, The Netherlands

Anne is a Process mining enthusiast since she came across this topic as a student. Process mining is a new business process management discipline that analyzes existing processes directly based on log data from the supporting IT systems. This analysis is done in a “bottom-up” way: The real processes are reconstructed starting from the data to support fact-based process improvements. Process mining answers questions such as: “How are our processes really executed?”, “Where are the bottlenecks?“, and “When and why do people deviate?”. No additional software needs to be installed to add this capability, because most IT systems already capture the historic information that is needed to perform process mining analysis. At Fluxicon, they have more than 15 years experience with Process Mining. They provide professional Process Mining software and trainings.

Email: anne@fluxicon.com
Web: https://fluxicon.com/team/

Lecture: Process Mining: Discovering process models based on data
Process mining is still relatively young, especially compared to other data analysis technologies like statistics or data mining, which have been around for several decades. At first, it looks deceivingly simple: You import a data set and the process mining tool automatically, almost magically, constructs a process map that shows you how the process was actually performed.

But in fact there are a lot of things that you need to know to get it right:
- What kind of data is suitable for process mining?
- How can you detect data quality problems?
- How do you interpret the results from the process mining tool?
- And what kind of questions can you answer with process mining in the first place?

Process mining is not just a tool but a new discipline that requires a smart human being who can make the connection between the data and the underlying business processes—with the help of the process mining tool.

In this lecture, we give you an overview about the typical process mining use cases but also about the challenges that you encounter when applying process mining techniques in practice.

Frederic Stahl

University of Reading, UK

Frederic Stahl completed his degree at the University of Applied Science in Weihenstephan (Germany) in Bioinformatics. He received the academic grade as Diploma Engineer in 2006 and obtained his Ph.D. from the University of Portsmouth in 2010. The title of his Thesis is “Parallel Rule Induction”. After his Ph.D. Frederic continued working as Senior Research Associate at the University of Portsmouth until February 2012. Frederic joined Bournemouth University as fixed term Lecturer from February 2012 until November 2012 and is currently working as Associate Professor at the University of Reading. His research interests are in the area of data mining of large and complex datasets; parallel and distributed data mining; data stream mining; data mining in resource constraint environments, machine learning and artificial intelligence.

Email: f.t.stahl@reading.ac.uk
Web: https://fredericstahl.wordpress.com/

Lecture: Exploring Algorithms and Approaches for the Real-Time Analytics of Data Streams
Developments in sensor and data acquisition technology and the increase of applications that require processing and analysing a constant, infinite stream of data records challenge our current data mining algorithms. Such applications are for example Internet of Things, Mirco-Blog rule mining (i.e. from Twitter), Sensor Networks, etc. This tutorial introduces the field of Data Stream Mining, relevant algorithms and includes some hands on tasks. Topics covered are:

* Data Stream Concept Drift Detection
* Predictive Data Stream Analytics
* Cluster Analysis on Data Streams

Practical tasks will require the prior installation of the Massive Online Analysis Software Release 2017.06 (https://moa.cms.waikato.ac.nz/downloads/).

Rossano Schifanella

University of Turin, Italy

Rossano Schifanella is an Assistant Professor in Computer Science at the University of Turin, Italy, where he is a member of the Applied Research on Computational Complex Systems group. He is a visiting scientist at Nokia Bell Labs and a former visiting scientist at Yahoo Labs and at the Center for Complex Networks and Systems Research at the Indiana University where he was applying computational methods to model the behavior of (groups of) individuals and their interactions on social media platforms. His research embraces the creative energy of a range of disciplines across data mining, network analysis, urban informatics, computational social science, and data visualization.

Email: schifane@di.unito.it
Web: http://www.di.unito.it/~schifane/

Lecture: Decoding Social Interactions in the Web
Due to the increasing availability of large-scale data on human behavior collected on the social web, as well as advances in analyzing larger and larger data sets, interdisciplinary research is working on building a better understanding of human societies bridging social science and computational methods. In this talk, we will explore classic social network analysis concepts (e.g., estimation of tie strength, extraction of communities), and we will present methods to interpret the role of microscopic and mesoscopic social structures (links and groups) in social interactions at scale. The talk will combine theoretical parts with a hands-on session in which we will show how to apply different methods and how to interpret results in a real case scenario of large-scale online social platform.

Niall Twomey

University of Bristol, UK

Niall received his PhD from University College Cork (Ireland) in 2013 for his research on the automated and remote detection of allergy as inferred from ECG signals. He is now a postdoctoral researcher in SPHERE project in the University of Bristol and his principal area of research currently involves data mining and fusion of environmental sensors in smart home environments. Niall’s personal research interests involve the use of digital signal processing, machine learning, data mining, and application-centric decision making for objective health and wellness assessment.

Email: Niall.twomey@bristol.ac.uk
Web: http://www.irc-sphere.ac.uk/uob-niall

Lecture: to be published soon

Marc Plantevit

Université Claude Bernard Lyon 1, France

Marc Plantevit received his PhD in computer science in 2008 from the University of Montpellier. He has been an associate professor in the computer science department of the University of Lyon since 2009. His research interests include constraint-based pattern mining in general. Currently, he is very interested with sophisticate pattern domains (dynamic/ attributed graphs) and in incorporating background knowledge into pattern mining.

Email: marc.plantevit@liris.cnrs.fr
Web: http://liris.cnrs.fr/~mplantev/doku/doku.php

Lecture: Preference-based pattern mining
This talk focuses on the recent shift from constraint-based pattern mining to preference-based pattern mining and interactive pattern mining. Constraint-based pattern mining, which shares common notions with FCA, is now a mature domain of data mining that makes it possible to handle various different pattern domains (e.g., itemsets, sequences, graphs) with a large variety of constraints thanks to solid theoretical foundations and an efficient algorithmic machinery. Even though, it has been realized for a long time that it is difficult for the end-user to model her interest in term of constraints and above to overcome the well-known thresholding issue, researchers have only recently intensified their study of methods for finding high-quality patterns according to the user’s preferences. In this lecture, we discuss the need of preferences in pattern mining, the principles and methods of the use of preferences in pattern mining. Many methods are derived from constraint-based pattern mining by integrating utility functions or interestingness measures as quantitative preference model. This approach transforms pattern mining in an optimization problem guided by user specified preferences. However, in practice, the user has only a vague idea of what useful patterns could be. The recent research field of interactive pattern mining relies on the automatic acquisition of these preferences and the development of the instant data mining field.

Alexandru C. Telea

University of Groningen, The Netherlands

Prof. Alexandru C. Telea has obtained his PhD in 2000 in data visualization and software architectures at the University of Eindhoven, the Netherlands. He has worked as assistant professor in visualization and computer graphics at the same university until 2007, when he has been appointed professor in software visualization at the University of Groningen, the Netherlands. His interests cover data, information, and software visualization, and static source code analysis and C/C++ reverse engineering. He has published over 130 papers in international peer-reviewed venues and one textbook in data visualization. He served as co-chair of IEEE EuroVis 2008, IEEE Vissoft 2007, ACM Softvis 2007 and general chair of ACM Softvis 2009. He is recognized as one of the main experts in the field of software visualization. His research has a strongly applied flavor, and he has co-developed several innovative tools and techniques for software visualization in the industry. He is a member of the ACM.

Email: a.c.telea@rug.nl
Web: http://www.cs.rug.nl/~alext/

Lecture: Visual Analytics Opens the Black Box of Machine Learning
In recent years, machine learning, and in particular deep learning, have made tremendous progresses both on theoretical and practical levels. Currently, machine learning algorithms are used by wide groups of practitioners, ranging from computer scientists, researchers in other exact science branches, technical people in data-intensive industries, and the grand public. However, with its successes, machine learning also has an Achille's heel: The operation of its algorithms is often opaque, and thus hard to understand, predict, and fine-tune. This raises serious problems for the effectiveness, replicability, trustworthiness, and ultimately applicability of machine learning in practical problems.

In this talk, we will discuss a broad range of techniques from information visualization and visual analytics that can be used to open this black box of machine learning for its interested users. Thereby, the general operation, unstable cases, learning behavior, and potential challenges encountered by machine learning techniques become visible, controllable, and actionable upon by its users. We will address questions such as: How to quickly determine if a classification problem is hard? How to understand what a neural network has learned (or not)? How to see which parts of some input data made a neural network take a given decision? And, above all, why do we need to understand deep learning black boxes for our businesses? Examples from real-world usage of visual analytics to help machine learning are presented coming from multiple application domains (image analysis, medicine, software engineering, and physics).

Martijn Willemsen

Eindhoven University of Technology, The Netherlands

Martijn Willemsen is associate professor in the Human-Technology Interaction group at the school of Innovation Sciences at the Eindhoven University of Technology and also works as the principle investigator of the Recommender Lab at Jheronimous Academy of Data Science in Den Bosch.

Martijn Willemsen researches the cognitive aspects of Human-Technology Interaction, with a strong focus on judgment and decision making in online environments. His applied research focuses on how online decisions can be supported by recommender systems, and includes domains such as movies, health related decisions and energy-saving measures. From a more theoretical perspective, he has a special interest in process tracing technologies to capture and analyze in detail information processing of decision makers.

Email: M.C.Willemsen@tue.nl
Web: http://www.martijnwillemsen.nl/

Lecture: Objective or Subjective measures for evaluation of recommender systems?
Recommender algorithms are traditionally evaluated using offline evaluation on historical datasets. More recently, focus has shifted to online evaluation of objective behavioral data using AB testing. However, such behavior is hard to interpret without using subjective measures that help interpreting the meaning of the behavior. For example lower click-rates might not be reflecting reduced interest, but increased engagement of a user consuming the recommended content from beginning to end without additional interactions. In this lecture I first introduce our user-centric evaluation framework (Knijnenburg & Willemsen, 2015) and subsequently show in several cases from prior work how objective (based on behavioral data) and subjective (based on user-surveys) measures go hand in hand in predicting and understanding user behavior and system effectiveness. As part of the lecture I will also provide a short tutorial on how to best perform user studies to test your systems.

Roxanne Arnts

SAP, The Netherlands

Roxanne Arnts is a machine learning advisor at Innovation Lab SAP, the Netherlands. She has a Bachelor of Business Administration (BBA) in Facility Management and a Master of Science (MSc) in Communication and Information Sciences from Tilburg University, The Netherlands.

Email: roxanne.arnts@sap.com
Web: https://www.linkedin.com/in/roxanne-arnts-15b027a8/

Lecture: Business Impact of Machine Learning
During the session we will have a look at how SAP, as a software provider, responds to the demand for Intelligent solutions. Learn about how different organizations are starting to see the added value of Machine Learning, and how they are taking the first steps into becoming an intelligent Enterprise.