Dissertation topics for PhD. study programmes for the academic year 2026/2027

Pavol Jozef Šafárik University in Košice, Faculty of Science

Institute of Computer Science

Study programme

Informatics (Id)

Title

Digital evidence analysis using machine learning methods

Language

SK

Abstract

Digital forensic analysis has become an essential component of incident response to computer security incidents as well as of cybercrime investigations. A key phase of a forensic investigation is the analysis of digital evidence. Within this phase, it is necessary to extract forensic artifacts, determine their relevance and evidential value for a given case, and identify relationships among them. The purpose of this phase is to confirm or refute forensic hypotheses formulated in the initial stages of the investigation.

Objective

(1) Analyze the possibilities of using machine learning methods for digital trace analysis with regard to the complexity, volume, and heterogeneity of forensic artifacts. (2) Propose an approach for selecting relevant forensic artifacts and discovering relationships among them. (3) Propose a method for verifying the forensic hypothesis itself.

References

(1) Hall, Stuart W., Amin Sakzad, and Kim‐Kwang Raymond Choo. "Explainable artificial intelligence for digital forensics." Wiley Interdisciplinary Reviews: Forensic Science 4.2 (2022): e1434. (2) Mohammad, Rami Mustafa A., and Mohammed Alqahtani. "A comparison of machine learning techniques for file system forensics analysis." Journal of Information Security and Applications 46 (2019): 53-61. (3) Tallón-Ballesteros, Antonio J., and José C. Riquelme. "Data mining methods applied to a digital forensics task for supervised machine learning." Computational intelligence in digital forensics: forensic investigation and applications (2014): 413-428. (4) Du, Xiaoyu, et al. "SoK: Exploring the state of the art and the future potential of artificial intelligence in digital forensic investigation." Proceedings of the 15th International Conference on Availability, Reliability and Security. 2020.

Tutor

doc. RNDr. JUDr. Pavol Sokol, PhD. et PhD.


Study programme

Informatics (Id)

Title

Analysis of tabular data using selected machine learning methods

Language

SK

Tutor

doc. RNDr. Ľubomír Antoni, PhD.

Consultant

prof. RNDr. Stanislav Krajči, PhD.


Study programme

Informatics (IdAj)

Title

Brain-training games for spatial hearing

Language

EN

Objective

Solutions designed to enhance auditory processing when hearing thresholds are within normal limits are very limited and none are as recognized or as widely available as are hearing aids and cochlear implants. The project aims to contribute to the development of novel procedures to rehabilitate auditory processing deficits (APD) by developing a brain training game based on modern auditory neuroscience and the results of the EU Horizon Europe SAV grant. The development of auditory brain training game will be in collaboration with Northeastern University Brain Game Center and Oregon Health State University. The main goal of the games is to develop and test rehabilitative techniques that restore auditory function for those who perform poorly on tests of APD by training various aspects of auditory processing.

References

Klingel M, Laback B, Kopco N (2021) Reweighting of Binaural Localization Cues Induced by Lateralization Training. Journal of the Association for Research in Otolaryngology, 22, 551–566, https://doi.org/10.1007/s10162-021-00800-8. Spisak O, Klingel M, Loksa P, Sebena R, Laback B, Kopco N (2021) “Spectral and binaural cue reweighting for sound localization in real and virtual environments,” 2nd Joint Conference on Binaural and Spatial Hearing, 7-8 October 2021.

Tutor

doc. Ing. Norbert Kopčo, PhD., univerzitný profesor


Study programme

Informatics (Id)

Title

Brain-training games for spatial hearing

Language

SK

Objective

Solutions designed to enhance auditory processing when hearing thresholds are within normal limits are very limited and none are as recognized or as widely available as are hearing aids and cochlear implants. The project aims to contribute to the development of novel procedures to rehabilitate auditory processing deficits (APD) by developing a brain training game based on modern auditory neuroscience and the results of the EU Horizon Europe SAV grant. The development of auditory brain training game will be in collaboration with Northeastern University Brain Game Center and Oregon Health State University. The main goal of the games is to develop and test rehabilitative techniques that restore auditory function for those who perform poorly on tests of APD by training various aspects of auditory processing.

References

Klingel M, Laback B, Kopco N (2021) Reweighting of Binaural Localization Cues Induced by Lateralization Training. Journal of the Association for Research in Otolaryngology, 22, 551–566, https://doi.org/10.1007/s10162-021-00800-8. Spisak O, Klingel M, Loksa P, Sebena R, Laback B, Kopco N (2021) “Spectral and binaural cue reweighting for sound localization in real and virtual environments,” 2nd Joint Conference on Binaural and Spatial Hearing, 7-8 October 2021.

Tutor

doc. Ing. Norbert Kopčo, PhD., univerzitný profesor


Study programme

Informatics (Id)

Title

Complexity aspects of automata and formal languages

Language

EN

Abstract

In selected models of automata, we examine descriptional complexity of various language operations with additional requirements, such as membership in a specific language class or an upper bound on alphabet size. We also consider the computational complexity of some decision problems related to formal languages and automata.

Objective

(1) Examine the operational complexity of concatenation on unary alternating finite automata. (2) Bridge the gap between the upper and lower bounds on the state complexity of the k-th power on suffix-free, prefix-closed, and suffix-closed languages. (3) Consider and solve some variants of the magic number problem for several regular operations, with respect to subclasses of regular languages.

References

(1) J.E. Hopcroft, J.D. Ullman, Introduction to Automata Theory, Languages and Computation, Addison-Wesley (1979) (2) M. Hospodár et al.: Descriptional complexity of the forever operator. Int. J. Found. Comput. Sci. 30(1), 115-134 (2019) https://doi.org/10.1142/S0129054119400069 (3) M. Hospodár et al.: Operational complexity: NFA-to-DFA trade-off. Inf. Comput. 307, 105369 (2025) https://doi.org/10.1016/j.ic.2025.105369 (4) A.N. Maslov: Estimates of the number of states of finite automata. Soviet Math. Doklady 11, 1373-1375 (1970) (5) M.O. Rabin, D. Scott: Finite automata and their decision problems. IBM Journal of Research & Development 3, 114-125 (1959) https://doi.org/10.1147/rd.32.0114 (6) S. Yu et al.: The state complexities of some basic operations on regular languages. Theoret. Comput. Sci. 125(2), 315-328 (1994) https://doi.org/10.1016/0304-3975(92)00011-F

Tutor

Ing. Michal Hospodár, PhD.


Study programme

Informatics (IdAj)

Title

Cross-modal interactions and spatial auditory processing

Language

EN

Objective

Vision influences how we perceive space by hearing. Ventriloquism effect and after-effect are phenomena illustrating short-term plasticity in spatial hearing induced by visual signals. Visual attentional cuing also influences spatial auditory processing both in terms of sound localization and spatial benefit in speech perception. The current project will examine the effect of visual information on spatial auditory perception by performing behavioral experiments, neuroimaging studies, and computational modeling.

References

Hladek L, Seitz A, Kopco N (2021) Auditory-visual interactions in egocentric distance perception: Ventriloquism effect and aftereffect. Journal of the Acoustical Society of America, 150, 3593-3607, doi.org/10.1121/10.0007066. Kopčo N, Lokša P, Lin I-F, Groh J, Shinn-Cunningham B (2019). Hemisphere-Specific Properties of the Ventriloquism Aftereffect. Journal of the Acoustical Society of America, 146, EL177 doi.org/10.1121/1.5123176

Tutor

doc. Ing. Norbert Kopčo, PhD., univerzitný profesor


Study programme

Informatics (Id)

Title

Forensic analysis of artificial intelligence systems

Language

SK

Abstract

Artificial intelligence systems are becoming an integral part of everyday life, which simultaneously leads to a significant increase in cyber threats and cybersecurity incidents. An important aspect of investigating such incidents is an adequate forensic investigation. Within this investigation, several challenges can be identified that are associated with the heterogeneity of available components that make up artificial intelligence systems.

Objective

(1) Analyze possibilities for the acquisition and extraction of digital evidence from artificial intelligence systems. (2) Analyze possibilities for using machine learning methods in the analysis of digital evidence from artificial intelligence systems. (3) Propose an automated approach for the extraction and analysis of forensic artifacts from artificial intelligence systems.

References

(1) Schneider, J., & Breitinger, F. (2023). Towards AI forensics: Did the artificial intelligence system do it?. Journal of Information Security and Applications, 76, 103517. (2) Vassilev, A., Oprea, A., Fordyce, A., & Andersen, H. (2024). Adversarial machine learning: A taxonomy and terminology of attacks and mitigations (NIST AI 100-5). National Institute of Standards and Technology. https://doi.org/10.6028/NIST.AI.100-2. (3) Jeong, D. (2020). Artificial intelligence security threat, crime, and forensics: Taxonomy and open issues. IEEE Access, 8, 184560-184574.

Tutor

doc. RNDr. JUDr. Pavol Sokol, PhD. et PhD.


Study programme

Informatics (Id)

Title

Fuzzy cellular automata

Language

SK

Abstract

The topic of the work is the investigation of fuzzifications of cellular automata, especially Conway's game of Life. In its classic version, it involves repeated discrete applications of a simple rule that adjusts the binary value of each cell of a square grid according to the values ​​of neighboring cells. The goal is to modify the rules of this game so that cells can also have non-binary values ​​(representing their "health") and to look for configurations that exhibit interesting behavior (for example, periodicity or other regularity) with these rules. Of course, the creation of the appropriate mathematical model and the ideal of the formulation and formal proofs of the obtained results are required.

Objective

The topic of the work is the investigation of fuzzifications of cellular automata, especially Conway's game of Life. In its classic version, it involves repeated discrete applications of a simple rule that adjusts the binary value of each cell of a square grid according to the values ​​of neighboring cells. The goal is to modify the rules of this game so that cells can also have non-binary values ​​(representing their "health") and to look for configurations that exhibit interesting behavior (for example, periodicity or other regularity) with these rules. Of course, the creation of the appropriate mathematical model and the ideal of the formulation and formal proofs of the obtained results are required.

Tutor

prof. RNDr. Stanislav Krajči, PhD.


Study programme

Informatics (Id)

Title

Fuzzy formal concept analysis

Language

SK

Abstract

Formal concept analysis is a data-mining method applied to a rectangular matrix of data in which each row corresponds to some object, each column corresponds to some possible attribute, and the matrix field value denotes a membership of the column attribute for row object. One of the goals of this method is to find so-called concepts, which are stable (in some sense) pairs of subsets of objects and attributes. The method can be considered a nice application of the algebraic notion of a Galois connection. It has been described in detail by Ganter and Wille, in particular for the so-called crisp case with binary matrix data. A natural question that arises is what happens if the matrix data are non-binary...

Objective

Formal concept analysis is a data-mining method applied to a rectangular matrix of data in which each row corresponds to some object, each column corresponds to some possible attribute, and the matrix field value denotes a membership of the column attribute for row object. One of the goals of this method is to find so-called concepts, which are stable (in some sense) pairs of subsets of objects and attributes. The method can be considered a nice application of the algebraic notion of a Galois connection. It has been described in detail by Ganter and Wille, in particular for the so-called crisp case with binary matrix data. A natural question that arises is what happens if the matrix data are non-binary...

Tutor

prof. RNDr. Stanislav Krajči, PhD.


Study programme

Informatics (Id)

Title

Cross-modal interactions and spatial auditory processing

Language

SK

Objective

Vision influences how we perceive space by hearing. Ventriloquism effect and after-effect are phenomena illustrating short-term plasticity in spatial hearing induced by visual signals. Visual attentional cuing also influences spatial auditory processing both in terms of sound localization and spatial benefit in speech perception. The current project will examine the effect of visual information on spatial auditory perception by performing behavioral experiments, neuroimaging studies, and computational modeling.

References

Hladek L, Seitz A, Kopco N (2021) Auditory-visual interactions in egocentric distance perception: Ventriloquism effect and aftereffect. Journal of the Acoustical Society of America, 150, 3593-3607, doi.org/10.1121/10.0007066. Kopčo N, Lokša P, Lin I-F, Groh J, Shinn-Cunningham B (2019). Hemisphere-Specific Properties of the Ventriloquism Aftereffect. Journal of the Acoustical Society of America, 146, EL177 doi.org/10.1121/1.5123176

Tutor

doc. Ing. Norbert Kopčo, PhD., univerzitný profesor


Study programme

Informatics (Id)

Title

Quantum structures and Formal concept analysis

Language

SK

Tutor

doc. RNDr. Ondrej Krídlo, PhD.


Study programme

Informatics (Id)

Title

Localization and extraction of structured data from texts

Language

SK

Tutor

RNDr. Peter Gurský, PhD., univerzitný docent


Study programme

Informatics (Id)

Title

Advanced neural network methods and their applications

Language

SK

Abstract

Neural networks are among the most popular machine learning methods and currently represent the state of the art in multimodal data analysis. Contemporary and impactful approaches include transformers, variational autoencoders, and diffusion models, which are applied across various domains, including computer vision, natural language processing, and synthetic data generation.

Objective

The aim of the doctoral thesis is to provide a comprehensive overview of current methods and to design and implement advanced neural networks and related architectures for computer vision tasks and other application domains. Specifically, the objective is to analyze existing models and their architectures, to design, implement, and experimentally evaluate a modified architecture with an emphasis on training stability and computational complexity, and to optimize the models for specific applications.

References

1. HE, Chunming, et al. Diffusion models in low-level vision: A survey. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2025. 2. RUSSELL, Stuart; NORVIG, Peter. Artificial Intelligence: a modern approach, 4th US ed., University of California, Berkeley, 2021 3. KELLEHER, John D. Deep learning. MIT press, 2019.

Tutor

doc. RNDr. Ľubomír Antoni, PhD.


Study programme

Informatics (Id)

Title

RAG system for legal texts

Language

SK

Tutor

RNDr. Peter Gurský, PhD., univerzitný docent


Study programme

Informatics (IdAj)

Title

Training of plasticity and attention in spatial hearing

Language

EN

Objective

In everyday situations, humans are exposed to multiple concurrent stimuli in complex, continuously changing environments. To correctly extract relevant information, they adapt their processing to reflect the specifics of the current scene, and they learn from previous experience to improve the perceptual strategies used. The current project proposes to perform a series of behavioral experiments, brain imaging studies, and computational modeling to study how training of attention and of mechanisms of implicit and explicit learning can be used to cope with complex listening environments for speech processing, sound localization, and learning of new phonetic categories.

References

Vlahou E, Ueno K, Shinn-Cunningham B, Kopco N (2021) Calibration of consonant perception to room reverberation. Journal of Speech, Language, and Hearing Research, 64(8), 2956-2976 Klingel M, Laback B, Kopco N (2021) Reweighting of Binaural Localization Cues Induced by Lateralization Training. Journal of the Association for Research in Otolaryngology, 22, 551–566, https://doi.org/10.1007/s10162-021-00800-8.

Tutor

doc. Ing. Norbert Kopčo, PhD., univerzitný profesor


Study programme

Informatics (Id)

Title

Training of plasticity and attention in spatial hearing

Language

SK

Objective

In everyday situations, humans are exposed to multiple concurrent stimuli in complex, continuously changing environments. To correctly extract relevant information, they adapt their processing to reflect the specifics of the current scene, and they learn from previous experience to improve the perceptual strategies used. The current project proposes to perform a series of behavioral experiments, brain imaging studies, and computational modeling to study how training of attention and of mechanisms of implicit and explicit learning can be used to cope with complex listening environments for speech processing, sound localization, and learning of new phonetic categories.

References

Vlahou E, Ueno K, Shinn-Cunningham B, Kopco N (2021) Calibration of consonant perception to room reverberation. Journal of Speech, Language, and Hearing Research, 64(8), 2956-2976 Klingel M, Laback B, Kopco N (2021) Reweighting of Binaural Localization Cues Induced by Lateralization Training. Journal of the Association for Research in Otolaryngology, 22, 551–566, https://doi.org/10.1007/s10162-021-00800-8.

Tutor

doc. Ing. Norbert Kopčo, PhD., univerzitný profesor


Study programme

Informatics (Id)

Title

Attacks on machine learning methods in cybersecurity

Language

SK

Abstract

Machine learning methods play an important role in incident response. For detecting security incidents or attacks, these methods learn models of normal behavior from training data and then identify incidents or attacks as deviations from that model. This process encourages attackers to manipulate training data so that the learned model fails to detect subsequent attacks. Beyond the training phase, security systems that use machine learning methods are also vulnerable to various attacks during the decision-making (inference) phase. An attacker can bypass the learned behavior of the detection system by using specially crafted inputs.

Objective

(1) Analyze machine learning methods used in cybersecurity with respect to their robustness against attacks. (2) Propose an approach for testing machine learning methods with regard to the possibility of adversarial misuse. (3) Propose methods for protecting machine learning techniques against various types of attacks.

References

(1) Debicha, Islam, et al. "Adv-Bot: Realistic Adversarial Botnet Attacks against Network Intrusion Detection Systems." Computers & Security (2023): 103176. (2) Pawlicki, Marek, Michał Choraś, and Rafał Kozik. "Defending network intrusion detection systems against adversarial evasion attacks." Future Generation Computer Systems 110 (2020): 148-154. (3) Richards, Luke E., Edward Raff, and Cynthia Matuszek. "Measuring Equality in Machine Learning Security Defenses." arXiv preprint arXiv:2302.08973 (2023).

Tutor

doc. RNDr. JUDr. Pavol Sokol, PhD. et PhD.