TUTORIAL on DATA MINING
from
SOCIAL, KNOWLEDGE, and SENSOR NETWORKS

Babović Zoran

Innovation Center of the University of Belgrade

Ph.D. Student

Bajec Marko

University of Ljubljana

Professor

Drašković Dražen

University of Belgrade

Ph.D. Student

Filipović Viktorija

University of Belgrade

Ph.D. Student

Filipović Vladimir

University of Belgrade

Professor

Furlan Bojan

University of Belgrade

Ph.D. Student

Jelisavčić Vladisav

Mathematical institute, Belgrade

Ph.D. Student

Kartelj Aleksandar

University of Belgrade

Ph.D. Student

Komlenac Damjan

University of Belgrade

Ph.D. Student

Mihajlović Aleksandar

Mathematical institute, Belgrade

Ph.D. Student

Milutinović Veljko

University of Belgrade

Fellow of the IEEE, Member of the Academia Europaea

Popović Aleksandra

Philips, USA

Project Director

Protić Jelica

University of Belgrade

Professor

Rakočević Goran

Mathematical Institute, Belgrade

Ph.D. Student

Tomić Mihailo

University of Belgrade

Ph.D. Student

The tutorial consist of the following parts:

(a) Mining from Social Networks
(b) Mining from the Internet
(c) Mining from the Medical Databases
(d) Mining from the Sensor Networks
(e) Mind Mining
(f) Implementations

The first section focuses on the state of the art in inteligent question routing systems and personality classification.

The second section presents an overview of the machine learning algorithms applied to Internet search
such as genetic algorithms and probabilistic topic models.

The third section presents some of the novel research in datamining from medical databases
performed at University of Belgrade in cooperation with some of the most significant institutions from the field
(part of the FP7 project ArtTreat).

The fourth section presents some of the novel research and applications
performed at the Inovation Center of the University of Belgrade
in the field of data mining from wireless sensor networks (part of the FP7 project ProSense).

The fifth section provides an insight into some of the topics designated for future research: Mindgenomics and Neuroeconomy.

The sixth section presents an overview of a dataflow supercomputing framework
used to accelerate datamining algorithms: Maxeler.