Overview
Learn how to overcome common challenges in topic modeling during this 13-minute conference talk from PyBay 2024. Explore an innovative approach that frames topic modeling as a search problem, enabling more effective real-time applications. Discover solutions to persistent issues like continually changing topics, stale document tags, and the difficulty of creating human-interpretable labels in unsupervised topic models. Gain insights into implementing both supervised and unsupervised topic models more effectively in production environments through this presentation by Kas Stohr at SF Python's PyBay conference.
Syllabus
Thinking of Topic Modeling as Search — Kas Stohr (PyBay 2024)
Taught by
SF Python