Seminar

Our seminar


Laboratory seminars (PV273 in the course catalog) on Wednesday, 10:30 – 11:30, A505, FI MU, Botanická 68, followed by an informal group lunch

The format of standard lectures: 30-40 minutes presentation + 15 minutes for questions, slides in English, presentation in English or Czech, based on the audience

    • 11.9. – 13.9.2025
      SitSem seminar at Telč
      See the program here
    • 1.10.2025 Canceled due to illness 
    • 8.10.2025
      RNDr. Terézia Slanináková, Jakub Čillík
      Rethinking search: interfaces, understanding, and navigating complexity
      Abstract: Search has become one of the most widely used digital technologies. Since Google’s original PageRank algorithm, search engines have evolved into hybrid systems that combine exact keyword matching with semantic embedding models, and, most recently, integration of LLMs. At the same time, user behavior has shifted as people increasingly rely on AI-powered assistants and conversational interfaces to find information.
      This talk surveys existing search technologies across industry and academia, identifying the data and the use cases they serve. It then explores two themes: (1) how to think about the requirements and trade-offs of a search system + how to decide what mode of understanding to choose (exact matching, semantic similarity, or reasoning on top of ontologies), and (2) how different search interfaces mediate between backend complexity and human users.
    • 15.10.2025
      doc. Mgr. Radek Pelánek, Ph.D., Mgr. Anna Řechtáčková
      Code quality defects in introductory programming 
      Abstract: We present our research on code quality in the context of introductory programming. We define the notion of a code quality defect and introduce our catalog of over 100 specific defects. We will demonstrate the EduLint tool (https://edulint.com/), share results from a survey of teachers at FI on the perceived severity of different defects, and report findings from our study of defect prevalence in student code. We will also outline ongoing research on automatically detecting poor identifier names using large language models (LLMs). The talk will conclude with an open discussion, and we look forward to your perspectives on the topic of code quality.
    • 22.10.2025
      prof. Ing. Vojtěch Spiwok, Ph.D.
      Language models in structure, dynamics, and design of proteins
      Abstract: Protein language models provide new opportunities to study protein structure and dynamics and to design new proteins. Here, we present our new approaches to address these topics.
      In natural language, words can have different meanings depending on the context of the whole text. This fact was successfully addressed by the introduction of transformer architecture into language translation, text generation, chatbots, or other applications. Inspired by this, we analyzed human proteins using a protein language model (ESM) at the level of individual amino acids. The aim is to make a map of amino acids in the context of proteins. Each amino acid is a point on the map. Close points represent amino acids in similar contexts. This will help to analyze functions of proteins and the discovery of new functions for proteins for which the function is unknown.
      Protein language models also provide information about dynamics of proteins. We tried to decipher the dynamics of the selected protein by altering the language model parameters. This will provide a computationally cheap way to characterize drug-binding sites in proteins.
      Finally, protein language models can be used to design new proteins. We will present our extension of the existing protein design method by parallel tempering and metadynamics.
    • 29.10.2025
      doc. RNDr. Radka Svobodová, Ph.D.
      From student to researcher: Exploring the essence of dissertability
      Abstract: We all intuitively understand that a student must undergo a certain development and intellectual transformation in order to become a doctor (and a proud holder of a Ph.D. degree). A doctorate enables a person to work independently in science.
      How can a student become a doctor in traditional scientific fields such as natural sciences, history, theology, mathematics, etc.? And how does informatics address this issue? And what is dissertability? I will reflect on these questions in my presentation.
    • 5.11.2025
      Romana Ďuráčiová
      TBA
    • 12.11.2025
      Discussion about the use of AI and LLM
      TBA
    • 19.11.2025
      RNDr. Martin Husák, Ph.D.
      Network-wide situational awareness: From network scanning to knowledge graphs
      Abstract: The lecture explores the concept of network-wide situational awareness, a critical capability for maintaining security, performance, and operational insight in modern computer networks. The session covers foundational and advanced techniques used to gain comprehensive visibility across an entire network environment. Key topics include network scanning and monitoring, which enable real-time detection of hosts and services; device fingerprinting, which involves identifying and classifying devices based on traffic patterns and behavioral signatures; and the construction of knowledge graphs, which integrate heterogeneous network data into a structured, queryable representation of assets and relationships. The knowledge graphs and underlying ontologies are emerging research trends that allow the network operators to represent holistic information on the network and, in the future, use LLMs to query the details on the network in natural language, generate reports and report issues, or even streamline autonomous network configuration.​
    • 26.11.2025
      RNDr. Terézia Slanináková
      PhD thesis rehearsal
    • 3.12.2025
      Mgr. Adrián Rošinec & his students
      TBA
    • 10.12.2025
      RNDr. Viktória Spišaková
      PhD thesis rehearsal
    • 17.12.2025 15:30
      Christmas Sitola

Past seminars

Contact: Hana Rudová

(c) 2021 SITOLA, Administration