Section outline

  • General

    Collapse all Expand all

  • Schedule and instructors

    Schedule

    • Monday, h 11:00 - 13:00 (Lab I)
    • Wednesday, h 9:00 - 11:00 (Lab H)

    Contacts

    • Prof. Giulio Rossetti | CNR-ISTI | giulio.rossetti@isti.cnr.it
    • Prof. Barbara Guidi | Università di Pisa | barbara.guidi@unipi.it

    Telegram group

  • Introduction

    Delving deep into the intricacies of complex systems—be they social, biological, or technological—is vital for accurately modeling and effectively tackling the pressing issues that define our world today. Take, for instance, the need to diminish polarization and radicalization in online discussions, predict high-frequency financial transactions or understand the dynamic interactions among proteins.

    Building upon the theoretical groundwork laid in "Social Network Analysis," this laboratory course aims to equip students with the tools to construct robust pipelines for analyzing complex systems derived from real-world data and represented as graphs. It will introduce key methodologies for data collection and preprocessing for graph analysis, delve into models designed to enhance the graph's descriptive capabilities and practical applications and explore techniques for evaluating experimental outcomes. The course emphasizes hands-on learning, empowering students to gain practical insights into these concepts through the utilization of advanced Python libraries.

    Exam

    The assessment of the course consists of

    • A group project (code & article-like report) to be carried out (partially) during classes with tutor’s support;
    • An oral discussion (focused on the project) aimed to evaluate the knowledge acquired.
    Grading
    • Project & Report: 70%
    • Presentation and Discussion: 30%

  • Textbooks and material

    Textbooks:

     

    Slides & Python Notebooks:

  • Lecture calendar

    Being a Laboratory course, lectures will take place in presence for MS students.
    PhD candidates are allowed to join remotely to through the MS Teams class.

    For each lecture are reported the book chapters/papers to be studied: while some of them are mandatory (i.e., most book chapters and surveys) others are suggested readings. 

    T: Theory lectures
    L:     Laboratory lectures
    G:     Group project labs

     
      Date Type Topic   Slides   Materials
     
     1

    		  
     

    		  T Intro & Network Science Recap
      Slides

    		  
         📕 Ch.2
     📙 Ch.3
     
     2

    		    T Where to start: Formulating Hypotheses
      Slides

    		  📗 Ch. 2-3
     3    G Group Project: group formation and available datasets

      Slides

    		  
     
     4

    		    T Modeling Choices: From simple graphs to advanced models   Slides   📙 Ch. 4
     
     5

    		    L Network Modeling: Static & Dynamic Networks

     Notebooks
    1, 2, 3, 4

    		  
     
     6

    		    L Network Modeling: Multiplex Networks  Notebook  
     
     7

    		    L Network Modeling: Higher-order structures  Notebook  
     8    T


    Enriching Topologies: Feature-rich modeling

    		  Slides  
     9    L


    Graph Augmentation: Node/Edge attributes

    		  Notebook  
     10    T


    Network Sampling

    		  Slides  📙 Ch. 25
     11    L Network sampling  Notebook   
     
     12    G Group project  

     13    T


    Data Collection: API, Web Scraping & Generation

    		  Slides  YSocial
     14    L


    Data Collection: API, Web Scraping & Generation

    		  Notebook  
     15    G


    Group project

    		    
     
     16

    		    T Graph Transformation  Slides  📙 Ch. 24
     17    L


    Graph Transformation

    		  Notebook  
     
     18

    		    G Group project    
     
     19

    		    G Group project    
     20    T


    How to Validate: check statistical significance of network-based studies

    		  Slides  📙 Ch. 16 
     📗 Ch. 4-5
     21    L


    Null model and statistical significance testing

    		    
     22    G


    Group project

    		    
     23    T/L


     Experiment reproducibility & Open Science

    		  Slides  
     24    T


     Experiments from A to Z: Examples and Case Study

    		    




  • Exercises

    Welcome to the exercises section!

    The activities in this section are designed to help you test your understanding of the theoretical concepts covered in lectures.

    • Formulating Hypotheses Assignment
      Opened: Friday, 18 September 2026, 9:49 AM

      The aim of this test is to develop your skills in formulating, evaluating, and refining scientific hypotheses, as well as distinguishing between well-constructed hypotheses and unscientific statements. You will also practice operationalizing hypotheses and linking them to appropriate research methods.

    • Modeling Choices Quiz
      Opened: Monday, 28 September 2026, 9:46 AM

      Answer all questions to evaluate your understanding of the theory covered during the lecture.

    • Feature-rich modeling Quiz

      Answer all questions to evaluate your understanding of the theory covered during the lecture

    • Graph Sampling Quiz

      Answer all questions to evaluate your understanding of the theory covered during the lecture.

  • Call For Thesis in Network Analysis

    Have you enjoyed the ACN course so much that you are considering a thesis on related subjects? Great!

    Here are a few ideas we would like to work on... of course you can also propose something new and completely different!

    • Mental Health in Online Social Media
    • Fake News, Echo Chamber, Polarization
    • Cognitive Network Science
    • Feature-rich network modeling
    • NDlib: modeling competing diffusion processes
    • NDlib: definition of novel diffusion models tailored for specific scenarios (e.g., fake news, opinion dynamics...)
    • CDlib: a comparative analysis of Community Discovery algorithms
    • CDlib: definition of novel community discovery approaches for dynamic/multiplex/feature-rich networks
    • XAI: Explaining Community Discovery Algorithms
    • DyNetX: a library for modeling and studying dynamic network topologies
    • Graph Embedding: representing (static|dynamic) networks in low dimensional space to support prediction and clustering
    • High Order Networks: studying and modeling high-order temporal networks

    A thesis can focus either on the definition of a new model/algorithm or on the usage of complex network analysis methodologies as tools for studying specific phenomena.

     

    Past Thesis example:

    • Data Warehouse For Reddit Data And Its Usage
    • Modeling peer-pressure: a data-driven estimate of open-mindedness from high-order online political discussions
    • Echo chamber and Political Polarization: A time- and linguistic-aware analysis of online polarized discussions
    • Towards Attributed Stream-Hypernetwork Analysis: Structure, Features and Dynamics of Complex Social Systems
    • Can language be modelled through Network Science? A Textual Forma Mentis Network case study
    • Filter Bubbles in the Italian Twittersphere: a data-driven analysis
    • A tale of two OSNs: online political debate around Capitol Hill assault on Twitter and Parler
    • Reasoning, Machine Learning and Network Science: a comprehensive approach to Knowledge Graphs interlinking
    • Fiori, reti e frattali: un approccio nature-inspired per l'analisi di network.
    • Scholarly Career Paths: Brain Drain and Exterophily
    • Polarizzazione Politica & Echo Chamber: una metodologia per l'identificazione e analisi su Reddit
    • Modeling Algorithmic Bias in Opinion Dynamics: simplicial complexes, and evolving network topologies
    • Modeling Human Mobility considering Spatial, Temporal and Social Dimensions
    • Bounded confidence, Stubborness e Peer Pressure: simulare l'adozione di fake news tramite modelli di opinion dynamics
    • La struttura complessa degli spazi semantici: un approccio guidato dalla network science
    • "It's a long way to the top" Predicting Success via Innovators'adoptions
    • Supervised Link Prediction su Reti Sociali

    Interested? Contact Giulio Rossetti and Barbara Guidi for more details.