What challenge does collaborative filtering face when new users are introduced?

The challenge that collaborative filtering faces when new users are introduced is primarily the data scarcity problem. This issue arises when a system lacks sufficient information about new users to make reliable recommendations. Collaborative filtering relies on user interactions, such as ratings or preferences, to identify patterns and suggest items. When a new user joins, the system does not have enough data about their tastes or preferences, making it difficult to draw meaningful comparisons with existing users.

This problem is often referred to as the "cold start" problem, wherein the lack of data prevents the algorithm from functioning effectively. As users interact with the system over time, their profile enriches, allowing the system to generate better recommendations. In essence, the efficacy of collaborative filtering hinges on the volume and richness of user data, which is initially lacking for newcomers.

While high computational costs, content availability, and complex algorithms can pose challenges in different contexts, they are not the fundamental issue when it comes to new users specifically. The heart of the challenge is rooted in the inability to gather sufficient interaction data to make accurate recommendations.