Feature Engineering

Organize feed features around the viewer, the author, the tweet, and the viewing context. For each candidate feature, explain what it measures and why it might help predict the target action. Treat the examples as hypotheses to test, not rules that define a person’s interests or the quality of their posts.

The machine learning model is required to predict user engagement on user A’s Twitter feed. Let’s engineer features to help the model make informed predictions.

Note: The feature set shown here is the result of one brainstorm session. However, feature engineering is an iterative process. As an exercise, you are encouraged to think about more features.

Let’s begin by identifying the four main actors in a twitter feed:

1. The logged-in user

2. The Tweet

3. Tweet’s author

4. The context

Features for the model