Data mining & algorithmic bias | GRTR256

This workshop will guide you through the first steps of understanding how to build machine learning models. No prior learning or background knowledge is necessary—students from all majors are encouraged to attend.

Before getting into a hands-on exercise, we will present a brief introduction to AI and machine learning, as well as the notion of algorithmic bias: what it is, how it is introduced, and its repercussions.

By the end of the workshop, you will have applied a technique for analyzing data according to gender to reveal the impact of biases in machine learning models. In other words, we will start building a sexist robot and learn how to spot its biases.

If you’re interested, we encourage you to visit the Quick Things for Digital Knowledge on machine learning or artificial intelligence before the session (about 10-15 minutes of reading/viewing).

NOTE

Computers will be provided at the session. However, if you want to use your own laptop, please install the open-source software Weka and download the data before the workshop.

Weka: https://www.cs.waikato.ac.nz/ml/weka/downloading.html

Datasets: http://labs.library.concordia.ca/workshops/textmining/datasets.zip

Participant of this workshop will:

• Understand basic notions of AI: machine learning and predictive models.
• Understand what is algorithmic bias, how is introduced in machine learning models and its possible repercussions.
• Transform text into word vectors (Bag of Words Approach) as a technique to perform text-mining tasks.
• Create a model for sentiment prediction using a machine learning approach based on a training corpus of real-life textual data (Tripadvisor comments on hotels and restaurants).
• Evaluate the model and compare the performance with different gender biased training corpuses.

This workshop is given by Francisco Berrizbeitia. Eng, M.Sc, Developer at Concordia Library.