NLP Club

Weekly NLP study and hands-on work group.

From September 2022 through May 2023 I was leading a weekly study group on NLP. Through this, we gain hands-on experience with a variety of NLP-related tasks, such as experimenting with different vectorizations and transformations, feature extraction, text classification, topic modeling, and getting familiar with the most common frameworks and APIs, such as scikit-learn, tensorflow, spaCy, NLTK, and huggingface, to name but a few.

For instance, we showed that simple models can outperform neural nets in sentiment classification (notebook, kaggle contribution), and used both classic LDA (notebook) and the amazing BERTopic (notebook) for (dynamic) topic modeling. We also did classification with neural nets using tensorflow (notebook). We also prototyped a web-based text analysis application for internal use that will support analysts in gaining insight from open text answers in surveys.

If you have any questions or feedback, don’t hesitate to contact me.