0
Historical Language Models in Cryptanalysis: Case Studies on English and German
Béata Megyesi, Justyna Sikora, Filip Fornmark et al., 2023. Linköping Electronic Conference Proceedings.
Tests whether hill-climbing and simulated-annealing solvers for homophonic ciphers work better with language models built from historical, century-specific English and German texts than with large modern ones.
What it's used for here
The cryptanalysis survey cites it as the reason to score solver output with period Latin and Italian language models, including one for abbreviated Latin; two other surveys also cite it.
Cited by on this site 3 records
3 methods surveys
SRCCryptanalysis, linguistics, encoding procedures
Cite it
Megyesi, B., Sikora, J., Fornmark, F., Waldispühl, M., Kopal, N., & Mikhalev, V. (2023). Historical Language Models in Cryptanalysis: Case Studies on English and German. Linköping Electronic Conference Proceedings. https://doi.org/10.3384/ecp195701
@article{megyesi2023historical,
author = {Megyesi, Béata and Sikora, Justyna and Fornmark, Filip and Waldispühl, Michelle and Kopal, Nils and Mikhalev, Vasily},
title = {Historical Language Models in Cryptanalysis: Case Studies on English and German},
journal = {Linköping Electronic Conference Proceedings},
year = {2023},
doi = {10.3384/ecp195701},
url = {https://doi.org/10.3384/ecp195701}
}TY - JOUR TI - Historical Language Models in Cryptanalysis: Case Studies on English and German AU - Megyesi, Béata AU - Sikora, Justyna AU - Fornmark, Filip AU - Waldispühl, Michelle AU - Kopal, Nils AU - Mikhalev, Vasily PY - 2023 JO - Linköping Electronic Conference Proceedings DO - 10.3384/ecp195701 UR - https://doi.org/10.3384/ecp195701 ER -
Megyesi, Béata et al.. “Historical Language Models in Cryptanalysis: Case Studies on English and German.” Linköping Electronic Conference Proceedings (2023). https://doi.org/10.3384/ecp195701.