Inhalt Hotkeys
Fakultät für Informatik


274. Informatik-Kolloquium


Freitag, 27. Oktober 2017, Straße der Nationen 62, Böttcher-Bau, 1/336

08:15 Uhr

Herr PD Dr. habil. Markus Endres

A tour of personalized data management

The advent of the Big Data era has created an explosion in the available data and information. In particular, user-generated content provided by people who (voluntarily) contribute data, e.g., hotel ratings, wiki articles, photos or videos, etc., has rapidly grown in recent years. However, to make this tremendous source of rich data useful, information retrieval and knowledge discovery methods are required to handle data from the heterogeneous sources. Personalization is one fundamental natural concept for information filtering and extraction to reduce the very large volume of data to a small set of highly interesting results. In my talk, I will address diverse concepts for personalized data management: I will give an overview of my recent research activities in this field by presenting challenges and solutions concerning preference handling in database systems.

10:00 Uhr

Herr PD Dr. Andreas Harth

Rule-based Data Integration and Service Composition with Linked Data-Fu

Linked Programs are designed for the declarative specification of applications that are composed of web-accessible components. Programs are based on rules that allow for the encoding of data access, deductive reasoning and data manipulation. Such programs can be evaluated with Linked Data-Fu, a general-purpose query and rule engine that operates a continuous sense-act cycle. Linked Data-Fu executes Linked Programs in a multi-threaded fashion and achieves an order of magnitude improvement in performance over a comparable engine. The talk introduces the interface to components based on web technologies, followed by a description of the system design of the rule engine. Further, real use-cases are presented to show the applicability of Linked Data-Fu in different scenarios.

Alle interessierten Personen sind herzlich eingeladen!