YouTube is one of the most popular websites for sharing videos, containing 60% of videos watched online. They use the tags assigned to a video as a means to improve content discovery and browsing of related content in a large collection of videos. We want to collect information regarding the textual properties of tags and their relationship with other metadata elements such as the title, description, and transcript.
We have a data set containing the metadata of 52,923 YouTube videos on the subject of "Java Programming". For 546 of these videos, we also have the English transcript available.
- This is structured like a single page app. All of the data is provided via the WebAPI, and AngularJS and Bootstrap are used to create the client interface.
- I intended for this project to run on local machines only. Thus, the interface excludes loading notfications and server error response like a published web application would.
- This is feature complete. The only planned changes are about tweaking how tags are defined.