Sebastian Baumbach

PhD Student
German Research Center for Artificial Intelligence, DFKI

Research Interests

Sensor Data Mining
Geopatial-temporal Data Mining
Decision Support Systems

Affiliation

Department of Smart Data & Knowledge Services

German Research Center for Artificial Intelligence, DFKI

Learn more about my research

My Research.

My main research area is Data Mining and Artificial Intelligence.

Research Interests

My main field of research is Data Mining. In particular, I am working on Sensor, Social and Spatio-temporal Data Mining.

Data-driven Site Selection

I am currently working on the QuIS Project, sponsored by the German Federal Ministry of Education and Research. The goal is to provide a data-driven approach for site selection of companies. In this context, I am working on spatial and temporal data analysis for recommendation and trend detection.

Teaching

I am working at the chair of Prof. Dengel at the Technical University of Kaiserslautern in the organization and teaching since 2014.

Here is my recent work.

Social Media Mining over Site Selection to Human Activity Recognition.

MyCustomer Logo

Sensor- & Media Mining for Participative Customer Relationship Management

In this project, we worked on a novel approach for analysing the interaction and activity behaviour of people. Therefore, their movements in daily life, so called trajectories, are gathered by modern smartphones. The objective was the identification of typical patterns, activities or groups in which people behave similarly or do similar things. Existing works do not take into account the various semantic information included in the trajectories. In addition to the pre-processing of people's movements and a spatio-temporal distance metrics, the novelty of this work is based on the hybrid clustering approach. Results have demonstrated that the method is very appropriate for applications such as optimization of traffic flow in cities, improvements of public transportation systems or analysing the purchasing behaviour.

Smartphone & Smartwatched Combined for Gathering Sensor Data

Supervising Sports Activities

The trend of training with a personal trainer is one of the most blooming trends in the recent years because of it's athletic benefits, however one cannot neglect its financial and time-wise compromises. The purpose of this project was to examine the ability of creating an automated sports personal trainer using the sensors integrated in the exerciser's smartphone and smartwatch to recognize the quality of execution of sports activities. As an approach to solve this problem, mobile and watch applications were developed to collect accelerometer, gyroscope and orientation sensors' data. Two data sets for user-dependent and user-independent configurations of the system were collected and tested for best recognition accuracy. Results showed that user-dependent archived very good results in terms of precision and thus, outperforms user-independent configurations.

Smartphone & Smartwatched Combined for Gathering Sensor Data

Human Activity Recognition

Unobtrusive and mobile activity monitoring using ubiquitous, cheap and widely available technology is the key requirement for human activity recognition supporting novel applications. These applications range from fitness monitoring, interactive game interfaces to home healthcare appliances. With the recent progress in wearable technology, pervasive sensing and computing has become feasible. However, recognizing complex activities on light-weight devices is a challenging task. In this work, methods for recognizing human activities in realtime using the sensors incorporated in smartphones and smartwatches were studied. However, the analysis results in this project showed that unobtrusive activity recognition by using ubiquitous devices like smartphones and smartwatches is feasible.

Publications.

Statistics

I have published 2 conference paper.

Details


2017

Measuring the Performance of Push-ups - Qualitative Sport Activity Recognition
Sebastian Baumbach, Andreas Dengel
Proceedings of the 9th International Conference on Agents and Artificial Intelligence (ICAART-17), Pages 374-381, Vol. 2, SCITEPRESS, Portugal, 2/2017, ISBN: 978-989-758-220-2
Test

2016

Human Activity Recognition Using Sensor Data of Smartphones and Smartwatches
Bishoy Sefen, Sebastian Baumbach, Andreas Dengel, Slim Abdennadher
Proceedings of the 8th International Conference on Agents and Artificial Intelligence (ICAART-16), Pages 488-493, Vol. 2, SCITEPRESS, Portugal, 2/2016, ISBN: 978-989-758-172-4
[paper]

Teaching.

Statistics

I have worked as Trainer for 5 lectures at Technische Universität Kaiserslautern. Additionally, I have given and organized 6 scientific seminars and projects.
During the last years, I have supervised 6 bachelor and 4 master thesis.

Details


Sommer 2017

Trainer: Lecture "Collaborative Intelligence", Masters's level

Organizer: Seminar "Artificial Intelligence", Bachelor's level

Master Thesis Supervisor: Analysing Spatio-Temporal Effects of Social-Economic Factors on Crime

Master Thesis Supervisor: Interactive Exploration and Visualization of Geospatial Temporal

Bachelor Thesis Supervisor: Geospatial Customer, Competitor, and Supplier Analysis for Site Selection

Winter 2016/17

Trainer: Lecture "Introduction to Symbolic Artificial Intelligence", Bachelor's level

Organizer & Topic Supervisor: Seminar "Collaborative Intelligence", Masters's level

Bachelor Thesis Supervisor: Erstellung eines Tools zur Visualisierung neuronaler Netze zur Klassifizierung linearer Zeitreihen

Sommer 2016

Trainer: Lecture "Collaborative Intelligence", Masters's level

Organizer: Seminar "Artificial Intelligence", Bachelor's level

Bachelor Thesis Supervisor: Incorporating Spatial Correlation between Sites into a Recommender System

Winter 2015/16

Trainer: Lecture "Introduction to Artificial Intelligence", Bachelor's level

Organizer & Topic Supervisor: Seminar "Collaborative Intelligence", Masters's level

Master Thesis Supervisor: Development and Integration of a Location Factor Model for Industrial Site Recommendation

Sommer 2015

Trainer: Lecture "Collaborative Intelligence", Masters's level

Organizer & Topic Supervisor: Seminar "Artificial Intelligence", Bachelor's level

Organizer & Project Supervisor: Project "Agile Methods II", Bachelor's level

Bachelor Thesis Supervisor: Towards Software Solution for Site Selection for Companies

Bachelor Thesis Supervisor: Supervising Sports Activities Using Sensor Data of Smartphones and Smartwatches

Winter 2014/15

Trainer: Lecture "Introduction to Artificial Intelligence", Bachelor's level

Organizer & Topic Supervisor: Seminar "Collaborative Intelligence", Masters's level

Master Thesis Supervisor: Human Activity Recognition Using Sensor Data of Smartphones and Smartwatches

Bachelor Thesis Supervisor: A Model for Aggregating Multi-layered, Spare and Geografically Referenced Data Sets

Contact.

Please feel free to contact me if you have any questions:

Sebastian Baumbach
Department of Smart Data & Knowledge Services
German Research Center for Artificial Intelligencey, DFKI GmbH
Room 3.33
Trippstadter Straße 122
D-67663 Kaiserslautern
Germany

E-mail: ed.ikfdsomething@hcabmuab.naitsabes

Phone: +49 631 20575 4250


Find me on ...