Integrating social networks properties such as centrality, tie strength, etc into message forwarding protocols in opportunistic networks has grown into a vital major benchmark. The opportunistic network is a demanding network with no set route to travel a message from the source to be able to the destination. During these networks, nodes use possibilities gained based on store-carry-forward patterns to forward communications. Every node that obtains a message when it activities another node makes selection concerning the forwarding or not necessarily delivering the node came across. Most of these message forwarding protocols use the benefit of social properties information like contact information and social relationship enclosed by the nodes in the social opportunistic network. In this paper, a Bonding based forwarding technique is proposed which is finding direct and indirect bonding among nodes by exploiting contact information and social pattern. In the proposed protocol, we also focus on indirect bonding by finding weakest direct bonded nodes and then replace it with strong indirect bonded nodes of the network. In this work, the balance between transmission delay and network traffic is considered by using shortest path map based mobility model. ONE simulator is used for simulation and performance of the proposed protocol is compared contrary popular approaches for instance Epidemic, PRoPHET, and BubbleRap, and Interaction based when using the shortest path map based mobility model. The Bonding based forwarding technique performs adequately well concerning the number of messages delivered, overhead ratio, message dropping and average latency.