Toward a Cross-Cultural and Cross-Language Multi-agent Recommendation Model for Food and Nutrition

In this paper, we present our efforts to develop a multi-agent-based framework for cross-cultural and cross-language personalized health and nutrition semantic search. Many agents that share either the same culture or same language or same health profile could share valuable information found through their semantic search. Their interests in a common viewpoint have some similarities which could be used to shorten the learning curve and give good search results effectively. Each agent is representing a user with a specific culture, language and personal health profile. What an agent likes as culture is likely similar to a different agent with the same culture. The same thing applies with language and health profiles. For example, an agent with a diabetes profile could help another agent that has a similar health profile with its valuable findings. We introduce how multi-agents could team up to contribute to each other, learn from each other, and share valuable information with each other using collaborative profile ontology. We propose a cross-cultural, cross-language, health- and nutrition-based ontology profile that could be used as the basis for collaboration between different agents.