Human-computer agent negotiation using cross culture reliability models

People’s cultural background has been shown to affect the way they reach agreements in negotiation and how they fulfill these agreements. This paper presents a novel methodology that can be used as a good infrastructure to design a computer-agent for negotiating with people from different cultures. Our setting involved data from different agents and human versus human data that were based on an alternatingoffer protocol that allowed parties to choose the extent to which they kept each of their agreements during the negotiation. A challenge to develop this methodology for such setting is to create cross culture models automatically that will predict how people reciprocate their actions over time, despite the scarcity of prior data on different cultures. Our methodology addresses this challenge by using a Leave-One-Out algorithm named CCMA, which is described in Sect. 5, with classical machine learning algorithms to predict the extent to which people fulfill agreements. Our methodology based its strategy on a data from different agents that used the same negotiation scenario in different cultures. This methodology used data in three countries: Lebanon, the U.S.A and Israel, in which people are known to vary widely in their negotiation behaviour. Our methodology was able to find the accurate models that should be used when designing a computer-agent in the negotiation scenario.