@misc{SerthPodlesnyBornsteinetal.2017,
  author    = {Serth, Sebastian and Podlesny, Nikolai and Bornstein, Marvin and Lindemann, Jan and Latt, Johanna and Selke, Jan and Schlosser, Rainer and Boissier, Martin and Uflacker, Matthias},
  title     = {An interactive platform to simulate dynamic pricing competition on online marketplaces},
  series = {2017 IEEE 21st International Enterprise Distributed Object Computing Conference (EDOC)},
  journal   = {2017 IEEE 21st International Enterprise Distributed Object Computing Conference (EDOC)},
  publisher = {Institute of Electrical and Electronics Engineers},
  address   = {New York},
  isbn      = {978-1-5090-3045-3},
  issn      = {2325-6354},
  doi       = {10.1109/EDOC.2017.17},
  pages     = {61 -- 66},
  year      = {2017},
  abstract  = {E-commerce marketplaces are highly dynamic with constant competition. While this competition is challenging for many merchants, it also provides plenty of opportunities, e.g., by allowing them to automatically adjust prices in order to react to changing market situations. For practitioners however, testing automated pricing strategies is time-consuming and potentially hazardously when done in production. Researchers, on the other side, struggle to study how pricing strategies interact under heavy competition. As a consequence, we built an open continuous time framework to simulate dynamic pricing competition called Price Wars. The microservice-based architecture provides a scalable platform for large competitions with dozens of merchants and a large random stream of consumers. Our platform stores each event in a distributed log. This allows to provide different performance measures enabling users to compare profit and revenue of various repricing strategies in real-time. For researchers, price trajectories are shown which ease evaluating mutual price reactions of competing strategies. Furthermore, merchants can access historical marketplace data and apply machine learning. By providing a set of customizable, artificial merchants, users can easily simulate both simple rule-based strategies as well as sophisticated data-driven strategies using demand learning to optimize their pricing strategies.},
  language  = {en}
}
@misc{PodlesnyKayemMeinel2019,
  author    = {Podlesny, Nikolai Jannik and Kayem, Anne V. D. M. and Meinel, Christoph},
  title     = {Attribute Compartmentation and Greedy UCC Discovery for High-Dimensional Data Anonymisation},
  series = {Proceedings of the Ninth ACM Conference on Data and Application Security and Privacy},
  journal   = {Proceedings of the Ninth ACM Conference on Data and Application Security and Privacy},
  publisher = {Association for Computing Machinery},
  address   = {New York},
  isbn      = {978-1-4503-6099-9},
  doi       = {10.1145/3292006.3300019},
  pages     = {109 -- 119},
  year      = {2019},
  abstract  = {High-dimensional data is particularly useful for data analytics research. In the healthcare domain, for instance, high-dimensional data analytics has been used successfully for drug discovery. Yet, in order to adhere to privacy legislation, data analytics service providers must guarantee anonymity for data owners. In the context of high-dimensional data, ensuring privacy is challenging because increased data dimensionality must be matched by an exponential growth in the size of the data to avoid sparse datasets. Syntactically, anonymising sparse datasets with methods that rely of statistical significance, makes obtaining sound and reliable results, a challenge. As such, strong privacy is only achievable at the cost of high information loss, rendering the data unusable for data analytics. In this paper, we make two contributions to addressing this problem from both the privacy and information loss perspectives. First, we show that by identifying dependencies between attribute subsets we can eliminate privacy violating attributes from the anonymised dataset. Second, to minimise information loss, we employ a greedy search algorithm to determine and eliminate maximal partial unique attribute combinations. Thus, one only needs to find the minimal set of identifying attributes to prevent re-identification. Experiments on a health cloud based on the SAP HANA platform using a semi-synthetic medical history dataset comprised of 109 attributes, demonstrate the effectiveness of our approach.},
  language  = {en}
}
@phdthesis{Podlesny2023,
  author    = {Podlesny, Nikolai Jannik},
  title     = {Quasi-identifier discovery to prevent privacy violating inferences in large high dimensional datasets},
  doi       = {10.25932/publishup-58784},
  url       = {http://nbn-resolving.de/urn:nbn:de:kobv:517-opus4-587843},
  school      = {Universit{\"a}t Potsdam},
  pages     = {xvi, 140},
  year      = {2023},
  abstract  = {Personal data privacy is considered to be a fundamental right. It forms a part of our highest ethical standards and is anchored in legislation and various best practices from the technical perspective. Yet, protecting against personal data exposure is a challenging problem from the perspective of generating privacy-preserving datasets to support machine learning and data mining operations. The issue is further compounded by the fact that devices such as consumer wearables and sensors track user behaviours on such a fine-grained level, thereby accelerating the formation of multi-attribute and large-scale high-dimensional datasets. In recent years, increasing news coverage regarding de-anonymisation incidents, including but not limited to the telecommunication, transportation, financial transaction, and healthcare sectors, have resulted in the exposure of sensitive private information. These incidents indicate that releasing privacy-preserving datasets requires serious consideration from the pre-processing perspective. A critical problem that appears in this regard is the time complexity issue in applying syntactic anonymisation methods, such as k-anonymity, l-diversity, or t-closeness to generating privacy-preserving data. Previous studies have shown that this problem is NP-hard. This thesis focuses on large high-dimensional datasets as an example of a special case of data that is characteristically challenging to anonymise using syntactic methods. In essence, large high-dimensional data contains a proportionately large number of attributes in proportion to the population of attribute values. Applying standard syntactic data anonymisation approaches to generating privacy-preserving data based on such methods results in high information-loss, thereby rendering the data useless for analytics operations or in low privacy due to inferences based on the data when information loss is minimised. We postulate that this problem can be resolved effectively by searching for and eliminating all the quasi-identifiers present in a high-dimensional dataset. Essentially, we quantify the privacy-preserving data sharing problem as the Find-QID problem. Further, we show that despite the complex nature of absolute privacy, the discovery of QID can be achieved reliably for large datasets. The risk of private data exposure through inferences can be circumvented, and both can be practicably achieved without the need for high-performance computers. For this purpose, we present, implement, and empirically assess both mathematical and engineering optimisation methods for a deterministic discovery of privacy-violating inferences. This includes a greedy search scheme by efficiently queuing QID candidates based on their tuple characteristics, projecting QIDs on Bayesian inferences, and countering Bayesian network's state-space-explosion with an aggregation strategy taken from multigrid context and vectorised GPU acceleration. Part of this work showcases magnitudes of processing acceleration, particularly in high dimensions. We even achieve near real-time runtime for currently impractical applications. At the same time, we demonstrate how such contributions could be abused to de-anonymise Kristine A. and Cameron R. in a public Twitter dataset addressing the US Presidential Election 2020. Finally, this work contributes, implements, and evaluates an extended and generalised version of the novel syntactic anonymisation methodology, attribute compartmentation. Attribute compartmentation promises sanitised datasets without remaining quasi-identifiers while minimising information loss. To prove its functionality in the real world, we partner with digital health experts to conduct a medical use case study. As part of the experiments, we illustrate that attribute compartmentation is suitable for everyday use and, as a positive side effect, even circumvents a common domain issue of base rate neglect.},
  language  = {en}
}
@misc{PodlesnyKayemvonSchorlemeretal.2018,
  author    = {Podlesny, Nikolai Jannik and Kayem, Anne V. D. M. and von Schorlemer, Stephan and Uflacker, Matthias},
  title     = {Minimising Information Loss on Anonymised High Dimensional Data with Greedy In-Memory Processing},
  series = {Database and Expert Systems Applications, DEXA 2018, PT I},
  volume    = {11029},
  journal   = {Database and Expert Systems Applications, DEXA 2018, PT I},
  publisher = {Springer},
  address   = {Cham},
  isbn      = {978-3-319-98809-2},
  issn      = {0302-9743},
  doi       = {10.1007/978-3-319-98809-2_6},
  pages     = {85 -- 100},
  year      = {2018},
  abstract  = {Minimising information loss on anonymised high dimensional data is important for data utility. Syntactic data anonymisation algorithms address this issue by generating datasets that are neither use-case specific nor dependent on runtime specifications. This results in anonymised datasets that can be re-used in different scenarios which is performance efficient. However, syntactic data anonymisation algorithms incur high information loss on high dimensional data, making the data unusable for analytics. In this paper, we propose an optimised exact quasi-identifier identification scheme, based on the notion of k-anonymity, to generate anonymised high dimensional datasets efficiently, and with low information loss. The optimised exact quasi-identifier identification scheme works by identifying and eliminating maximal partial unique column combination (mpUCC) attributes that endanger anonymity. By using in-memory processing to handle the attribute selection procedure, we significantly reduce the processing time required. We evaluated the effectiveness of our proposed approach with an enriched dataset drawn from multiple real-world data sources, and augmented with synthetic values generated in close alignment with the real-world data distributions. Our results indicate that in-memory processing drops attribute selection time for the mpUCC candidates from 400s to 100s, while significantly reducing information loss. In addition, we achieve a time complexity speed-up of O(3(n/3)) approximate to O(1.4422(n)).},
  language  = {en}
}