PUMA publications for /group/simtech/peerFri Mar 22 10:19:41 CET 2024Proceedings of the International Conference on Data Engineering (ICDE)FairCR - An Evaluation and Recommendation System for Fair Classification Algorithms2024myown intcdc peer rp9-2 peerReviewed ipvs-de accountability Mon Jan 22 10:40:10 CET 2024International Conference on Extending Database Technology (EDBT)FALCC: Efficiently performing locally fair and accurate classifications2024myown intcdc peer rp9-2 peerReviewed ipvs-de accountability Thu Oct 27 08:18:39 CEST 2022International Conference on Extending Database Technology (EDBT)80--91Progressive Entity Resolution over Incremental Data2023intcdc entity-resolution peer rp9-2 ipvs-de Thu Oct 27 08:12:19 CEST 2022Intelligent Information Systems - Forum, International Conference on Advanced Information Systems Engineering (CAiSE)45--54The ArchIBALD Data Integration Platform: Bridging Fragmented Processes in the Building Industry2022intcdc peer ipvs-de rp9 Wed Feb 09 14:17:23 CET 2022Datenbank-Spektrum123 - 43Metrics and Algorithms for Locally Fair and Accurate Classifications using Ensembles222022intcdc peer ipvs-de rp9 accountability Thu Jul 01 02:39:02 CEST 2021Proceedings of the VLDB Endowment (PVLDB)Debugging Missing Answers for Spark Queries over Nested Data with Breadcrumb2021pebble why-not peer ipvs-de rp9 Thu Jul 01 02:34:24 CEST 2021Procedia CIRP31st CIRP Design Conference 2021 (CIRP Design 2021)451-456Integrative data processing for cyber-physical off-site and on-site construction promoting co-design1002021peer ipvs-de rp9 The term co-design describes a collaborative design process, usually across the boundaries of individual areas. A major aspect, and often an obstacle in the implementation of the co-design approach, is the establishment of a centralized data management system. In the past, the data management concept has been approached from different perspectives. Current approaches usually are oriented along the process chain and thus reflect a unidirectional flow of information. Neither the information feedback nor the reaction to adjustments in upstream model steps is supported, which makes continuous collaboration more difficult and at best reduces the co-design approach for continuous data usage. This paper addresses these weaknesses while outlining an integrative data processing concept enabling co-design between the domains of construction and industrial prefabrication. Geometrical information generated during architectural design and construction planning is reused for the production of components, manufacturing-related design adjustments are transferred back into the domain of construction. Changes throughout the entire process are centrally recorded. To ensure independence from the software tools used and to consider their great diversity, data exchange is largely based on standards, namely industry foundation classes and the standard for the exchange of product model data. In order to counteract the sequential processing according to a waterfall model and to support the co-design approach, the industrial prefabrication toolchain, consisting of computer-aided manufacturing software and downstream postprocessors, is largely automated, so that reprocessing of a workpiece for minor geometric adjustments no longer requires any interaction of the production planner.Mon Mar 15 07:53:06 CET 2021In Proceedings of Database Systems for Business, Technology, and Web (BTW)Using FALCES against bias in automated decisions by integrating fairness in dynamic model ensembles2021intcdc peer ipvs-de rp9 accountability Thu Oct 22 14:34:20 CEST 2020Proceedings of the IEEE International Conference on Data Engineering (ICDE)End-to-end Task Based Parallelization for Entity Resolution on Dynamic Data2021entity-resolution peer ipvs-de rp9 Thu Mar 26 09:53:44 CET 2020In ACM International Conference on Management of Data (SIGMOD)1893-1896Capturing and Querying Structural Provenance in Spark with Pebble2019pebble peer d03 ipvs-de trr161 rp9