Notebooks for the Introduction to Human Language Technology Lab sessions.
Master courses: Research Master Humanities and Text Mining.
This GitHub contains the python notebooks for the course Introduction into Human Language Technology at the Vrije Universiteit Amsterdam, Faculty of Humanities. This course targets Language and AI Master students and Research Master Humanities students. We assume you have some basic knowledge of Python or are following the Python for NLP course that runs in parallel.
The notebooks are grouped in lab sessions:
- lab0: getting started and learn to work with Natural Language Processing toolkits
- lab1: chatting with Large Langauge Models and annotating the conversation with emotion labels
- lab2: modeling the meaning of words: wordnets and word embeddings
- lab3: modeling the meaning of text: machine learning for Natural Language Processing
- lab4: evaluating text interpretation systems
- lab5: large language models
- lab6: generative large language models
Lab7 contains the instructions for the final assignment that is graded.
During the first LAB session you are getting prepared to work with different tools and text data.
We want you to work on your own computer so that you can create your own code and freely experiment. A solid laptop and substantial disk space helps. We are going to use the following applications and toolkits:
- The Terminal under Linux or Mac OS, the command line under windows.
- Anaconda environment: https://anaconda.org, which also installs Python 3.10 or higher
- Jupyter notebooks: https://jupyter.org. (you also may want to check out this: https://towardsdatascience.com/bringing-the-best-out-of-jupyter-notebooks-for-data-science-f0871519ca29
To prepare for the course, make sure you have installed Anaconda and you familiarised yourself with the environment and with notebooks. Please follow the instruction in the document HLT-getting-started.pdf, which you find at the root of this repository.
We assume that you have basic programming skills in Python and that you can work in the Command Line from a terminal.
In addition to the above tools, you need a plain text editor to inspect text files that form your data. Note that this is not the same as Word or Adobe. Word and PDF files are binary files and contain many other things besides the text (just like HTML files on websites continue instructions how to visualise text). The following text editors can be used:
- Windows: Notepad++
- Mac OS/Linux: Atom, BBEdit
Tip: to find out the difference between proper text files and other types of document, open one of the above text editors and load a Word, Pdf or Html file. You will see immediately a lot of stuff that is not text.
