This document discusses the potentials and challenges of recommendation systems for software engineering. It covers:
1. There is a large need for recommendation systems to address the main challenges in software projects, such as daily changes, information overload, distributed development, and high competition.
2. Some initial research systems have shown promising early results but have limitations in the types of scenarios and contexts they support.
3. Future recommendation systems should support both knowledge consumers and providers during development and collaboration tasks.
4. Major research challenges include better modeling, representing, contextualizing and sharing knowledge in software engineering projects.