Bert meets BERT

Again, learning through doing. This time I decided to spice things up with some good old Sesame Street. I had been hearing a lot about BERT topic modeling, so I figured it was time to dive in.

What made this project especially fun was working with data sources I had not used before. The notebook pulls data from Wikimedia, YouTube transcripts, and a few other sources using existing APIs and packages. I honestly did not realize how easy it had become to pull this kind of data directly into a notebook and store it locally for analysis.

In the end, the topic modeling results from the Sesame Street data did not reveal anything particularly surprising. But that was not really the point. The real value was learning how the pipeline works, experimenting with the tools, and getting more comfortable working with these kinds of datasets. Definitely a fun project and a good way to learn.