Introduction to Social Media
Twitter Algorithm
It is important to know that responsible machine learning (ML) is a long journey requiring continuous improvement as machine learning (aka algorithms) can potentially impact hundreds of millions of Tweets per day. Twitter has also created a 'Machine Learning Ethics, Transparency, and Accountability (META) team' to continually evaluate the algorithm.
The Twitter algorithm focuses on personalization as it looks at these ranking signals:
- Tweet: the recency, presence of images or video posts, and total interactions
- Tweet's Author: past interactions with the author of a twee, the strength of connection, and the origin of the relationship
- Individual: the tweet's the individual found engaging in the past
It is key to understand that, for Twitter, timing has substantial weight in the algorithm, that credibility is important, and that likes and retweets will have a weighting score.
References:
- Machine Learning Initiative:
https://blog.twitter.com/topics/introducing-responsible-machine-learning-initiative - How the Twitter Algorithm Works:
https://blog.hootsuite.com/twitter-algorithm/ - Build an Advanced Twitter Strategy:
https://academy.hubspot.com/courses/twitter-strategy - Social Media:
https://en.wikipedia.org/wiki/Social_media - Social Media Marketing: Certification Course
https://academy.hubspot.com/courses/social-media