Abstract: Emergency response planning is a complex task due to multiple organizations involved, different planning considerations, etc. Using artificial intelligence collaborative planning helps in the automatic planning for complex situations. Analyzing all impacting factors along with plans that are executable can facilitate the decision making in Emergency Operations Centers for an agile emergency response. A main component of a planner is a knowledge base. Although many systems are developed to support decision making in emergency response or recovery, they either focus on specific or small organizations, or rely on simulations. To the best of our knowledge, there is a gap that there is no common knowledge base for provincial level mass emergencies for automatic planners. The multiplicity of the emergency response documents and their structure makes the knowledge acquisition complex. In this paper, we explain the process of extracting knowledge based on hierarchical task networks and how it speeds up the reactivity to a disaster.