Tom Hanika

My research is concerned with the foundations and applications of explainable artificial intelligence. I am particularly interested in the (interactive) extraction of knowledge from explicit and implicit data by means of machine learning procedures. At the very heart of my work is the study of semantics arising from implicational theories in complex and large data. Of special interest in this context are lattice orders and ordinal-metric spaces. Problems in high dimensions are of special interest to me, as they represent a current challenge for the effectiveness of learning procedures. Apart from that I am very interested in unsupervised learning from large text corpora as well as applications of my methods in various research fields, such as Biology, Geography, Physics, Digital Humanities, and more.

Besides my work in research I am also a senior developer of the social publication sharing system BibSonomy as well as maintainer of the FCA-tool conexp-clj.

selected publications

Tobias Hille, Maximilian Stubbemann, Tom Hanika

Reproducibility and Geometric Intrinsic Dimensionality: An Investigation on Graph Neural Network Research.

Transactions on Machine Learning Research (2024).

URL Bibtex

@article{hille2024reproducibility,
  title = {Reproducibility and Geometric Intrinsic Dimensionality: An Investigation on Graph Neural Network Research.},
  author = {Hille, Tobias and Stubbemann, Maximilian and Hanika, Tom},
  journal = {Transactions on Machine Learning Research},
  issn = {2835-8856},
  year = {2024},
  url = {https://openreview.net/forum?id=CtEGxIqtud},
  note = {Reproducibility Certification},
}

Johannes Hirth, Viktoria Horn, Gerd Stumme, Tom Hanika

Ordinal motifs in lattices

Information Sciences 659 (2024), 120009.

Abs doi Bibtex

Lattices are a commonly used structure for the representation and analysis of relational and ontological knowledge. In particular, the analysis of these requires a decomposition of a large and high-dimensional lattice into a set of understandably large parts. With the present work we propose /ordinal motifs/ as analytical units of meaning. We study these ordinal substructures (or standard scales) through order-embeddings and (full) scale-measures of formal contexts from the field of formal concept analysis. We show that the underlying decision problems are NP-complete and provide results on how one can incrementally identify ordinal motifs to save computational effort. Accompanying our theoretical results, we demonstrate how ordinal motifs can be leveraged to achieve textual explanations based on principles from human computer interaction.
@article{HIRTH2024120009, title = {Ordinal motifs in lattices}, journal = {Information Sciences}, volume = {659}, pages = {120009}, year = {2024}, month = jan, issn = {0020-0255}, doi = {10.1016/j.ins.2023.120009}, url = {https://www.sciencedirect.com/science/article/pii/S0020025523015943}, author = {Hirth, Johannes and Horn, Viktoria and Stumme, Gerd and Hanika, Tom}, keywords = {Ordered sets, Explanations, Formal concept analysis, Closure system, Conceptual structures}, }

Katharina B. Budde, Christian Rellstab, Myriam Heuertz, Felix Gugerli, Tom Hanika, Miguel Verdú, Juli G. Pausas, Santiago C. González-Martínez

Divergent selection in a Mediterranean pine on local spatial scales

Journal of Ecology 112 (2024), no. 2, 278-290.

doi Bibtex

@article{https://doi.org/10.1111/1365-2745.14231,
  author = {Budde, Katharina B. and Rellstab, Christian and Heuertz, Myriam and Gugerli, Felix and Hanika, Tom and Verdú, Miguel and Pausas, Juli G. and González-Martínez, Santiago C.},
  title = {Divergent selection in a Mediterranean pine on local
                    spatial scales},
  journal = {Journal of Ecology},
  keywords = {candidate genes, FST outlier, local adaptation,
                    paired sampling design, Pinus pinaster, south-
                    versus north-facing slopes},
  doi = {10.1111/1365-2745.14231},
  url = {https://besjournals.onlinelibrary.wiley.com/doi/abs/10.1111/1365-2745.14231},
  volume = {112},
  number = {2},
  pages = {278-290},
  eprint = {https://besjournals.onlinelibrary.wiley.com/doi/pdf/10.1111/1365-2745.14231},
  year = {2024},
}

Maximilian Stubbemann, Tom Hanika, Friedrich Martin Schneider

Intrinsic Dimension for Large-Scale Geometric Learning

Transactions on Machine Learning Research (2023).

URL Bibtex

@article{stubbemann22,
  added-at = {2022-10-12T10:58:53.000+0200},
  author = {Stubbemann, Maximilian and Hanika, Tom and Schneider, Friedrich Martin},
  biburl = {https://www.bibsonomy.org/bibtex/222e3a51007f8947640b29167996fe48c/tomhanika},
  description = {Intrinsic Dimension for Large-Scale Geometric
                    Learning},
  interhash = {ddb49d07cf62c8b138a8a223d4753081},
  intrahash = {22e3a51007f8947640b29167996fe48c},
  journal = {Transactions on Machine Learning Research},
  keywords = {journal,selected},
  timestamp = {2023-01-13T13:27:54.000+0100},
  title = {Intrinsic Dimension for Large-Scale Geometric
                    Learning},
  url = {https://openreview.net/forum?id=85BfDdYMBY},
  year = {2023},
}

Bastian Schäfermeier, Johannes Hirth, Tom Hanika

Research Topic Flows in Co-Authorship Networks

Scientometrics 128 (2023), no. 9, 5051–5078.

doi Bibtex

@article{schafermeier2022research,
  author = {Schäfermeier, Bastian and Hirth, Johannes and Hanika, Tom},
  biburl = {https://www.bibsonomy.org/bibtex/23688d07e1cd8318ef9665336665825b5/tomhanika},
  description = {Research Topic Flows in Co-Authorship Networks},
  doi = {10.1007/s11192-022-04529-w},
  interhash = {6e8cb91056c87bd63bb2046be8055cea},
  intrahash = {3688d07e1cd8318ef9665336665825b5},
  issn = {1588-2861},
  journal = {Scientometrics},
  keywords = {journal},
  month = sep,
  number = {9},
  pages = {5051--5078 },
  timestamp = {2023-08-13T14:43:50.000+0200},
  title = {Research Topic Flows in Co-Authorship Networks},
  volume = {128},
  year = {2023},
}

Bernhard Ganter, Tom Hanika, Johannes Hirth

Scaling Dimension

In: 17th ICFCA, Proceedings, D. Dürrschnabel, D. López-Rodríguez (eds.), 64–77. Springer (2023).

doi Bibtex

@inproceedings{DBLP:conf/icfca/GanterHH23,
  added-at = {2023-08-02T16:54:41.000+0200},
  author = {Ganter, Bernhard and Hanika, Tom and Hirth, Johannes},
  bibsource = {dblp computer science bibliography,
                    https://dblp.org},
  biburl = {https://www.bibsonomy.org/bibtex/2ae364c1012bb099c906b24ecdc7f688d/tomhanika},
  booktitle = {17th {ICFCA}, Proceedings},
  doi = {10.1007/978-3-031-35949-1\_5},
  editor = {Dürrschnabel, Dominik and López-Rodríguez, Domingo},
  interhash = {a3f20563676f3181056c601d1def547a},
  intrahash = {ae364c1012bb099c906b24ecdc7f688d},
  keywords = {conf},
  pages = {64--77},
  publisher = {Springer},
  series = {Lecture Notes in Computer Science},
  timestamp = {2023-08-02T16:54:41.000+0200},
  title = {Scaling Dimension},
  url = {https://doi.org/10.1007/978-3-031-35949-1\_5},
  volume = {13934},
  month = jul,
  year = {2023},
}

Gerd Stumme, Dominik Dürrschnabel, Tom Hanika

Towards Ordinal Data Science

Transactions on Graph Data and Knowledge 1 (2023), no. 1, 6:1–6:39.

doi Bibtex

@article{DBLP:journals/tgdk/StummeDH23,
  author = {Stumme, Gerd and Dürrschnabel, Dominik and Hanika, Tom},
  title = {Towards Ordinal Data Science},
  journal = {Transactions on Graph Data and Knowledge},
  volume = {1},
  number = {1},
  pages = {6:1--6:39},
  year = {2023},
  url = {https://doi.org/10.4230/TGDK.1.1.6},
  doi = {10.4230/TGDK.1.1.6},
  timestamp = {Tue, 19 Dec 2023 17:15:12 +0100},
  biburl = {https://dblp.org/rec/journals/tgdk/StummeDH23.bib},
  bibsource = {dblp computer science bibliography,
                    https://dblp.org},
  keywords = {journal},
}

Tom Hanika, Friedrich Martin Schneider, Gerd Stumme

Intrinsic dimension of geometric data sets

Tohoku Mathematical Journal 74 (2022), no. 1, 23 – 52.

doi Bibtex

@article{intrinsic,
  added-at = {2022-01-30T18:56:40.000+0100},
  author = {Hanika, Tom and Schneider, Friedrich Martin and Stumme, Gerd},
  biburl = {https://www.bibsonomy.org/bibtex/234cd317c997745ba26b40684b702987a/tomhanika},
  doi = {10.2748/tmj.20201015a},
  interhash = {3b9cbb6abf829a466684a2d55f7da4b7},
  intrahash = {34cd317c997745ba26b40684b702987a},
  journal = {Tohoku Mathematical Journal},
  keywords = {journal,selected},
  number = {1},
  pages = {23 -- 52},
  publisher = {Tohoku University, Mathematical Institute},
  timestamp = {2022-01-30T18:56:40.000+0100},
  title = {Intrinsic dimension of geometric data sets},
  volume = {74},
  year = {2022},
  issn = {0040-8735},
}

Tom Hanika, Johannes Hirth

On the lattice of conceptual measurements

Information Sciences 613 (2022), 453-468.

doi Bibtex

@article{conmeasure,
  author = {Hanika, Tom and Hirth, Johannes},
  biburl = {https://www.bibsonomy.org/bibtex/2752bc4553487e6472b22faf56e744cad/tomhanika},
  doi = {10.1016/j.ins.2022.09.005},
  interhash = {ecd4e48ac83c8aea9197f7e3d861b032},
  intrahash = {752bc4553487e6472b22faf56e744cad},
  issn = {0020-0255},
  journal = {Information Sciences},
  keywords = {journal,selected},
  pages = {453-468},
  timestamp = {2022-09-28T00:14:02.000+0200},
  title = {On the lattice of conceptual measurements},
  url = {https://www.sciencedirect.com/science/article/pii/S0020025522010489},
  volume = {613},
  year = {2022},
}

Daniel Borchmann, Tom Hanika, Sergei Obiedkov

Probably approximately correct learning of Horn envelopes from queries

Discrete Applied Mathematics 273 (2020), 30 - 42.

doi Bibtex

@article{pacexp,
  author = {Borchmann, Daniel and Hanika, Tom and Obiedkov, Sergei},
  biburl = {https://www.bibsonomy.org/bibtex/2be121e13826f0ad3699d2629f08cc77f/tomhanika},
  doi = {10.1016/j.dam.2019.02.036},
  issn = {0166-218X},
  journal = {Discrete Applied Mathematics},
  keywords = {journal,selected},
  pages = {30 - 42},
  timestamp = {2020-02-21T11:59:05.000+0100},
  title = {Probably approximately correct learning of Horn
                    envelopes from queries},
  url = {http://www.sciencedirect.com/science/article/pii/S0166218X19301295},
  volume = {273},
  year = {2020},
}