MethodsNet + RSS
MethodsNet + RSS

Advanced text analysis (RSS1.18) - Closed

Classical content analysis used to be an expensive (monetarily and computationally) method. Nevertheless, with the constant technological advancement, the availability of data is at an all-time high, and the methods to analyze the data are constantly increasing in potential. At its core, text-as-data approaches have the same aim as classical content analysis – extracting meaning out of text. However, due to the unstructured and multidimensional nature of texts, there are additional challenges in achieving this goal. 

Duration: one-week.

    General


    This course is closed, registration is no longer possible. 

    Registration for this course is no longer possible yet we would like to offer you some similar courses for which registration is still possible:

    • ''Big Data: Introduction to Text Analysis,'' which covers some of the same topics as in the ''Advanced Text Analysis'' course. Feel free to contact the instructor (Bruno de Castanho Silva : b.paula.castanho.e.silva [at] fu-berlin.de (b[dot]paula[dot]castanho[dot]e[dot]silva[at]fu-berlin[dot]de)) for more information about the course, including what topics will be covered and the level of the course.
    • "Use of AI for Political Analysis", this course also focuses on large language models that would have also played a prominent role in the course ''Advanced Text Analysis". Feel free to contact the instructor (Ethan Busby: ethan.busby [at] byu.edu) for information about the course and the level of models covered in that course.  
    Collaboration

    Watch what our participants say about their experience!

    MethodsNet + RSS

    (Social) Research Methods courses

    Course list
    Course list

    Overview courses and disciplines

    Course list

    Cannot join us this year? 

    We can keep you informed about the 2025 course program! Do you want to broaden your knowledge in 2025 over courses about sustainability, law, research methods & skills, data science and more. Get an email when the new proposal is ready. Because you have part to play!

    Keep me informed

    Starting date

    17 June 2024, 9 am
    Educational method
    On-site
    Main Language
    English
    Sessions
    17 June 2024, 9 am - 21 June 2024, 5 pm
    Teacher(s)
    Petro Tolochko
    Unique code
    RSS1.18

    Factsheet

    Type of education
    Course
    Entry requirements
    See the requirements in cost and admission
    Study load (ECTS)
    2
    Result
    Certificate

    Course program

    This course is aimed at people who have some experience with text-as-data approaches, but what to understand more nuanced aspects of methods used to analyze texts. The topics that are covered in the course are: Text Representation (ways to transform unstructured text for computational analyses), Statistical models of texts, Word Embedding Models, and Deep Learning models. The advantages and disadvantages of each method/approach are also discussed. The topics are introduced in an interactive lecture-type setting, while the practical part consists of a coding session with examples and tasks. Small homework assignments are given out throughout the course week to deepen the knowledge of the topics.

    • Day 1: Motivation / Examples / Text Representation
    • Day 2: Statistical Models of language 
    • Day 3: Supervised / Unsupervised Machine Learning 
    • Day 4: Deep Learning / Word Embeddings 
    • Day 5: Transformer Models / Large Language Models

    Total package & social events

    Petro Tolochko

    Petro Tolochko
     

    Petro Tolochko is currently a postdoctoral researcher in the PolarVis and ySkills project. Before ySkills, he worked as a postdoctoral researcher at the Department of Political Science and the ERC project MARIPOLDATA.

    Petro completed his Master’s degree at the University of Amsterdam (ASCoR) and defended his doctoral thesis at the University of Vienna in June 2019. His doctoral thesis was concerned with the automated approached to determine the text complexity of political texts.

    His research interests include: quantitative methods in the social sciences, social networks, quantitative text analysis, and automated language processing in the social sciences.

    Costs  

    This course is closed, registration is no longer possible

    If you would like to stay informed for this course or similar courses for next year, please fill out our ''keep me informed'' form here

    • Regular: €1049 (application deadline 1st of May)
    • Student & PhD's: €699 (application deadline 1st of May)

    Includes: your course, short morning and late afternoon courses, coffee and tea during breaks, a warm lunch every day, Official Opening, MethodsNET Café (including some drinks and snacks) Official Closing (with some drinks and snacks) and a 1-year (2024 calendar year) free membership as MethodsNET regular member.

    Excludes: transport, accommodation, social events and other costs. 

     

    Admission

    Level of participant: 

    • Master
    • PhD
    • Postdoc
    • Professional

    Admission requirements: 

    Basic statistics (Bachelor level) - Familiarity with programming languages.

    Admission documents: 

    None