Welcome to the websites of the Databases and Information Systems Group (DBIS) at the department of computer science at TU Kaiserslautern.
We work on performance and quality aspects of handling Big Data, specifically for rendering large amounts of heterogeneous and evolving data accessible to users. This connects research on advanced data analytics and data exploration. We work on the development of novel algorithmic solutions, system architectures, and data indexing schemes, using and extending core database concepts as well as data mining, information retrieval, and machine learning techniques.
In the DFG-funded Pantheon project, we investigate means to assemble entity-centric rankings based on categorical as well as numerical facts reported in entities on the Web or in knowledge bases in particular. Another research branch deals with scaling out join computations for large-scale entity matching and in general theta joins. Read on under Research and Publications
You are studying Computer Science at TU Kaiserslautern and are searching for possible topics of Bachelor's or Master's theses? You are interested in data management and data analysis techniques and have successfully attended respective lectures? Then do not hesitate to contact us directly!
New full paper at ICDE 2022: "BETZE: Benchmarking Data Exploration Tools with (Almost) Zero Effort".
We are part of the BMEL funded DigiVine project: More information under https://digivine.org/
One full paper and two short papers accepted at EDBT 2020. On Distributed Similarity Joins, learned indices, and data wrangling, respectively.
Our work on "Scaling Out Schema-free Stream Joins" has been accepted to ICDE 2020 as a full paper, and additionally we come to present our demonstration "JODA: A Vertically Scalable, Lightweight JSON Processor for Big Data Transformations" in Dallas in April. Looking forward to seeing you there!
The demo proposal of our system CLASH got accepted at SIGMOD! CLASH is a high-level abstraction on top of Apache Storm, for expressing and optimizing multi-way stream joins. Check out the teaser video https://youtu.be/oZxNIwvEQDw
Our approach to identify interesting categorical attributes based on a classification approach using old and novel statistical features got accepted at SSDBM 2018!
Our brand-new approach (COMPETE) that harnesses dominance relationships between competing, ranked entities got accepted at ExploreDB@SIGMOD 2018. See you in Houston next month!
New paper on optimizing class-constraint nearest neighbor queries accepted at WebDB@SIGMOD 2018.
Our paper on pruning and querying inverted indices got accepted at WebDB 2017@SIGMOD and our work on scaling out multi-way theta joins got accepted at BeyondMR 2017@SIGMOD. See you in Chicago in May!
Good news: The German Research Foundation (DFG) has accepted our proposal to extend/prolong our research on entity rankings and data exploration within the Pantheon project.