This repository includes R code that can be used to assess the similarity between a YouTube video and the related videos recommened by YouTube, as defined by the YouTube API.
See these blog posts for a description of the project for which this code was developed:
https://sites.temple.edu/tudsc/2020/02/26/identifying-similar-youtube-videos-by-comparing-metadata-in-r/ https://sites.temple.edu/tudsc/2020/05/07/identifying-similar-youtube-videos-by-comparing-metadata-in-r-part-2/
The code runs in RStudio Cloud. You will need to upload the .csv of YouTube Categories in your working directory before running the code.
To run the code, you will need OAuth credentials to access the YouTube API. See Creating YouTube API . . . instructions.
You will also need to customize the dictionary with key terms from your chosen (initial) video.