Context awareness plays an important role in  Ubiquitous applications based on the context information to provide services adapting to users’ needs. The advantages of some context-aware smart phone systems, including iCAM[1], contextPhone[2], Live Contact Project [3], Enhanced Telephony[4] and Ubiquitous phone System Ubiphone [5] show themselves to be typical examples of Ubiquitous  service ones. 

Using a context-aware smart phone, we can contact individuals in many ways, including by home, office, or cell phone, short 

message service (SMS), instance message(IM), and email. Therefore, a human-centered smart phone could discover how 

best to reach a contact at any given moment. However, in these systems, the users can contact to people in the contact list only, 

and the context information is managed and referred on only one ontology tree in one domain which contraries to the reality that in Ubiquitous environment. In this environment, the context information such as concepts and relationships are also managed 

and referred on heterogeneous ontology in several domains, which drive new challenges in the discovery of knowledge sources relevant to a users’ request. New efficient technique and approach for developing and extending Ubiquitous phone 

intelligence is presented in this paper, which combines agent-based technologies and Artificial Neural Networks (ANNs) 

models to match ontology, helping users search and contact with anyone outside the contact list satisfying the users’ requests even there is no social relationship between them.