In this paper, we present an ontology-based approach to analyze depression of family caregivers of Alzheimer’s disease patients. First, we developed depression ontology called OntoDepression considering the language written in social media. Four major classes and specialized subclasses are defined based on the dailyStrength, which is a well-known social media site centered on healthcare. Next, to find mental health of family caregivers of Alzheimer’s patients, their twitter data is analyzed based on the OntoDepression. Our experimental results show that negative feelings of family caregivers are not clearly revealed, while medical condition of depression symptom is highly rated. Also, their tweets mention a lot about human relationships, work and activities.