Smart Proactive Assistance

Digital devices have become our daily companions. Companions which augment the human intellect just like Douglas Engelbart envisioned. Those companions offer access to information and services anywhere and anytime. Our research tackles the transformation of the digital device from a passive collection of apps to a personal assistant. Apple Siri, Google Now and Microsoft Cortana show the beginning of this transformation. Just like a real assistant the devices should know the user, should know the habits, the appointments, the community and the interests of the user. Based on such knowledge the assistant helps organizing ones live. 

The Smart Personal Assistance Area conducts research to shape the vision of personal assistance and to create methods for intention specific user support. To address this challenge, we focus on the human activity and involved information consumption. We work on the 1) deviation of knowledge from historic user data – habits, interests, etc. 2) information processing and aggregation – collect and connect data. To realize these goals we apply techniques from data mining and machine learning.

Current Research is a platform for the collection of activity-related data. serves as a quantified-self tool and helps users to understand their daily activities better. Furthermore, activity data collected with the platform can be used to do research in the domain of personal assistance.
Mobile Cloud Computing
The basic idea of mobile cloud computing is the usage of ambient computation resources/cloudlets to execute complex computations. We especially consider techniques for the dynamic identification of computation resources and the distribution of calculations on those devices. As a first showcase we use the routers as computation resources.
Predictive Maintenance for Cargo Trains
Trains are equipped with a large number of sensors. In this project we harness the sensor readings to predict maintenance events and failures. Doing so can reduce equipment downtime and aid in improving maintenance planning. As a result the operations efficiency improves and cost is reduced.
Self-Service Process Mining
Process Mining helps organizations to gather valuable information about the actual use of information systems. However, visual evaluation of discovered process models is hard because of the increasing amount of data. We do research to automatically mine insights using machine learning and data mining, reducing the amount of manual work currently required.

Former Projects

Student Assist
The student assistant project aims at supporting students during their university related activities. We develop a system which understands student activities and proactively supports organization of learning activities and the daily live. The tool applies techniques of activity recognition and advanced analysis techniques of locations and information access.
Advanced Content Mining
Goal of advanced content mining is the filtering and organization of content based on user preferences. Our activities include the analysis of social media (e.g., microblogs) to identify information and events relevant for the individual. We build consolidated models of social networks. Furthermore, we work on techniques to combine news data to improve the overall understanding news stories.

Additional Information

If you are interested in a Bachelor, Diploma, or Master thesis do not hesitate to contact us. We are also looking for motivated student workers (HiWis).


Sebastian Kauschke (Research Assistant)

Christian Meurisch (Research Assistant)

Timo Nolle (Research Assistant)

Alexander Seeliger (Research Assistant)

A A A | Drucken Print | Impressum Impressum | Sitemap Sitemap | Suche Search | Kontakt Contact | Webseitenanalyse: Mehr Informationen
zum Seitenanfangzum Seitenanfang