#StandWithUkraine 💛💙

Hi, ahoj, salut, hola, servus! 👋

As a sociologist at Charles University with a focus on media communication, I am interested in exploring how technology tools, such as Natural Language Processing and Machine Learning, can benefit my discipline and promote transparency and accountability in society. In addition to conducting research and implementing data science solutions, I am also passionate about teaching and sharing my knowledge with others. My goal is to make meaningful contributions to the field of social science and support the development of a healthy democratic system.

🔭 Experience:

• Languages: R & Python, operational knowledge of Bash, HTML/CSS/JS/Jamstack, SQL and regex.
• NLP & Linguistic methods: text classification, named entity recognition, collocations & concordances, word embeddings, sentiment analysis (lexicon & BERT model), semantic analysis (UCREL & UDPIPE features), topic modeling, KW extraction, document similarity.
• Statistical methods: sequence analysis, regression modeling, factor analysis, SEM (lavaan), clustering, explainable AI (SHAP).
• Data extraction: REST APIs, scraping automation using Selenium.
• Database software: DBeaver, Google BigQuery.
• Cloud computing: Jupyter-based environments (Kaggle, Deepnote, Colab, SageMaker Lab, Azure ML), Google Cloud Platform, AWS.
• Deployment & automation: GitHub Actions, Netlify.
• Containerization tools: Docker.
• Other relevant stack experience: RStudio, VS Code, Postman, Observable JS, Quarto, GraphQL, Power BI, Tableau, Julia.

🌱 Interested in:

• End-to-end open source data solutions.
• Applications of NLP & ML on political discourse corpora.
• Building automatic data pipelines.
• Advanced data analysis and visualizations.

👯 Projects:

• PhD-related research: analysis of ~1 mio. Czech news media articles on migration-related topics, 2015-2022 together with a wider social context.
• FAMU Prague: analysis of ~0.5 mio. transcripts of Czech public news media, 2012-2022, with particular focus on climate change reporting.
• Transparency International CZ: Several solutions for automated data extraction, analysis and visualizations of paid political advertising on Meta platforms and political party financing (on transparent bank accounts) in the context of 2023 presidential elections, 2022 municipal elections and 2021 parliamentary elections in the Czech Republic.
• Chapter in the open-source book "APIs for social scientists" on the usage of Facebook Ad Library API.
• Contributed to Facebook's Radlibrary package, which provides a high-level interface for FB Ad Library API in R.
• Covid pandemic, psychological health, and media sources: Analysis of an international survey dataset from the "Pandemic Emergency in Social Perspective" project.

🤔 I am looking for collaboration on:

• Web app to aid non-native language speakers with correct syntax using existing models with translation into auxiliary language and back.
• Object classification and visual sentiment analysis of social media images.
• Finalization of my open-source code into standalone R libraries (or PL improvements of original packages).

Among the many mysteries of the universe, I am still trying to figure out this age-old conundrum 🐶:

graph LR

  A(<br/><br/>Dog<br/><br/><br/>) --> B(Knocks over Water Glass)
  B --> C(Slips on Water)
  C --> D(Runs Away)
  D --> E(Hides Under Table)
  E --> F(Looks Innocent)
  F --> G(<br/><br/>Gets Fed<br/><br/><br/>)
  G ----> A
  style A fill:#e91e63, stroke:#000000
  style B fill:#9c27b0, stroke:#000000
  style C fill:#00bcd4, stroke:#000000
  style D fill:#4caf50, stroke:#000000
  style E fill:#ffeb3b, stroke:#000000
  style F fill:#ff6855, stroke:#000000
  style G fill:#1EE9A4, stroke:#000000