The task of planning humanitarian relief operations within a high number of hardly collaborating and vaguely linked nongovernmental organizations is a challenging problem. The authors suggest an alternative knowledge-based approach to the coalition formation problem for humanitarian and peacekeeping missions. Owing to the special nature of this domain, where the agents representing individual organizations may eventually agree to collaborate but are often reluctant to share their knowledge and resources, the authors tried to reduce the problem complexity by splitting the community of agents into alliances. They combined classical negotiation mechanisms with the acquaintance models and social knowledge techniques in order to reduce the communication traffic and to keep the privacy of knowledge.