ABC News Youtube Comment Sentiment Analysis
For a university assignment I scraped comments across several of Australia’s ABC News YouTube channel videos and performed sentiment analysis on said data. Firstly, I wrote a critical analysis paper in which I describe the dynamics of YouTube and other social media scraping, sentiment analysis and authorship attribution using academic research and other resources. This paper can be found here. Secondly, I wrote a paper based on the exploration of the scraped data and its sentiment (which was calculated using SentiStrength.) This includes looking in how users post across videos and some of the behaviours of top-posting users. This can be found here.
A Jupyter Notebook PDF showing the code for scraping, analysis and visualizations can be found here.
Takeaways
- Learnt how to navigate YouTube API for scraping.
- Able to dip into social media analysis research, with YouTube being rarely looked into.
- Pipeline included running the data through SentiStrength and analyzing and visualizing it with Python and Jupyter Notebook.
If I Had More Time…
I would have worked on authorship contribution in my research paper using my scraped data. However, this is a nascent field and would have taken me more time than what I had to research, comprehend, and apply to my data, with the possibility of very disappointing and unuseful results.
comments