Skip to content


Folders and files

Last commit message
Last commit date

Latest commit


Repository files navigation

Python Tutorials for NLP, ML, AI

(C) 2016-2024 by Damir Cavar

NLP-Lab at Indiana University.


NLTK Notebooks

spaCy Notebooks

See the licensing details on the individual documents and in the LICENSE file in the code folder.


The files in this folder are Jupyter-based tutorials for NLP, ML, AI in Python for classes I teach in Computational Linguistics, Natural Language Processing (NLP), Machine Learning (ML), and Artificial Intelligence (AI) at Indiana University.

If you find this material useful, please cite the author and source (that is Damir Cavar and all the sources cited in the relevant notebooks). Please let me know if you have some suggestions on how to correct the notebooks, improve them, or add some material and explanations.

The instructions below are somewhat outdated. I use just Jupyter-Lab now. Follow the instructions here to set it up on different machine types and operating systems.

To run this material in Jupyter you need to have Python 3.x and Jupyter installed. You can save yourself some trouble by using the Anaconda Python 3.x distribution.

Clone the project folder using:

git clone

Some of the notebooks may contain code that requires various kinds of [Python] modules to be installed in specific versions. Some of the installations might be complicated and problematic. I am working on a more detailed description of installation procedures and dependencies for each notebook. Stay tuned, this is coming soon.

Installing Jupyter

Jupyter is a great tool for computational publications, tutorials, and exercises. I set up my favorite components for Jupyter on Linux (for example Ubuntu) this way:

Assuming that I have some of the development tools installed, as for example gcc, make, etc., I install the packages python3-pip and python3-dev:

sudo apt install python3-pip python3-dev

After that I update the global system version of pip to the newest version:

sudo -H pip3 install -U pip

Then I install the newest Jupyter and Jupyterlab modules globally, updating any previously installed version:

sudo -H pip3 install -U jupyter jupyterlab

The module that we should not forget is plotly:

sudo -H pip3 install -U plotly

Scala, Clojure, and Groovy are extremely interesting languages as well, and I love working with Apache Spark, thus I install BeakerX as well. This requires two other [Python] modules: py4j and pandas. This presupposes that there is an existing Java JDK version 8 or newer already installed on the system. I install all the BeakerX related packages:

sudo -H pip3 install -U py4j
sudo -H pip3 install -U pandas
sudo -H pip3 install -U beakerx

To configure and install all BeakerX components I run:

sudo -H beakerx install

Some of the components I like to use require Node.js. On Ubuntu I usually add the newest Node.js as a PPA and not via Ubuntu Snap. Some instructions how to achieve that can be found here. To install Node.js on Ubuntu simply run:

sudo apt install nodejs

The following commands will add plugins and extensions to Jupyter globally:

sudo -H jupyter labextension install @jupyter-widgets/jupyterlab-manager
sudo -H jupyter labextension install @jupyterlab/plotly-extension
sudo -H jupyter labextension install beakerx-jupyterlab

Another useful package is Voilà, which allows you to turn Jupyter notebooks into standalone web applications. I install it using:

sudo -H pip3 install voila

Now the initial version of the platform is ready to go.

To start the Jupyter notebook viewer/editor on your local machine change into the notebooks folder within the cloned project folder and run the following command:

jupyter notebook

A browser window should open up that allows you full access to the notebooks.

Alternatively, check out the instructions how to launch JupyterLab, BeakerX, etc.