This repo contains an annotated jupyter notebook with all the relevant code used in the creation of my MSc. Thesis as well as the Thesis itself as PDF. My thesis marks the first time I tried coding in Python, start to finish is what taught me the basics. As can be seen throughout; none of the formal programming paradigms were used, some code duplication persists and many functions could have been improved for efficiency. Current day, my coding skills are much improved over what is seen here, nonetheless it was a fascinating project to gain my first experience through :). For questions about the content, please be referred to my LinkedIn.
Please note that directories and keys have been anonimized within the notebooks.

1. Sampling V1; Here we sample videos based on channel ids and select them for further analysis, using the YouTube API.
2. Sampling V2 (Omitted)
3. Statistics Retrieval; Here we use the YouTube API to find video statistics such as like count, age, etc.
4. Transcription; Here we first use youtube-dl to download audio tracks which we then transcribe using the IBM Watson Speech to Text api
5. Natural Language Processing; Here we define our features and process our data.

All of the above chapters are also compiled in one complete document, this is the last notebook in repo. The analysis portion of the thesis has not been included since it is completely tailored to my thesis but is available upon request.