How do you make recommendations for a brand-new user?
mediumAnswer
- Fall back through a ladder of decreasing personalization.
- With no history, serve popularity, ideally popularity within whatever context you do know, such as country, device, or referral source, which is far better than global top items.
- Use any content signal available: the first item viewed, a stated interest at signup, or the search query, and recommend by content similarity rather than collaborative signal.
- Deliberately explore in the first sessions, since early diverse impressions are how you acquire the data personalization needs, and a purely greedy policy starves itself.
- Then switch to the collaborative model once the user crosses an interaction threshold.
- A hybrid model that consumes content features handles this natively instead of needing a separate rule.
Check yourself — multiple choice
- Show random items
- Ladder down: contextual popularity, then content-based similarity from any available signal, with deliberate exploration, switching to collaborative filtering past an interaction threshold
- Wait until they have 100 interactions
- Cold start has no solution
Contextual popularity plus content signals and exploration bridge the gap until collaborative signal exists.
#applications
Practise Unsupervised Learning
214 interview questions in this topic.