Leonard Traeger, Ph.D. ’25, information systems, recently won the Best Paper Award at the International Conference on Enterprise Information Systems (ICEIS). The title of his paper is “Scoping: Towards Streamlined Entity Collections for Multi-Sourced Entity Resolution with Self-Supervised Agents.”
About his work Traeger states, ” My research motivation stems from the fact that ‘Data Scientists spend more time looking for data than analyzing it’ (Stonebraker 2018). Unfortunately, even with the help of AI, it remains a challenge to automatically find correct linkages across tables and attributes between two or more data systems due to quality and scalability constraints. With ‘Scoping’, we propose to use outlier algorithms to narrow down the search space of potential linkable entities and therefore increase the performance of the interlinked system at the same time. As simple as common outlier preprocessing methods such as Z-Score and Local-Outlier-Factor sound, applying these in the context of Data Integration boosts entity linkage algorithms due to the streamlined data entity collections. Beyond these baseline methods, we present and adapt novel encoder-decoder architectures. Entities passing such self-trained neural networks receive a linkability score based on the reconstruction error. We transform these scores and use them to schedule and reduce the number of similarity checks even further to enable a more concise system integration.”
Held this year at the end of April in Angers, France, the purpose of the ICEIS is to bring together researchers, engineers and practitioners interested in the advances and business applications of information systems, covering different aspects such as Enterprise Database Technology, Systems Integration, Artificial Intelligence, Decision Support Systems, Information Systems Analysis and Specification, Internet Computing, Electronic Commerce, Human-Computer Interaction and Enterprise Architecture.
Leonard Traeger is a Ph.D. candidate and research assistant working with Dr. George Karabatis. He received his bachelor’s (2019) and master’s (2022) degrees in Information Systems from the Technical University of Cologne, Germany. In Fall 2021, Leonard visited UMBC as a Fulbright Scholar where his research focused on improving and automating Data Integration between multiple systems and domains using Machine Learning approaches and Semantics. With his prior three year’s experience in consulting large-scale Data Warehouses in the public sector, his teaching philosophy and scientific contributions aim to be both innovative and practical in real-world applications.
Congratulations, Leo!