Sentiment based service selection in mobile ad hoc networks

The availability of the Web 2.0 applications in the mobile devices have enabled the users to express sentiments about their invoked services in fast and flexible ways. These sentiments, accessible over the Internet, can greatly influence a potential service user´s selection especially in a service rich ambiance. Often, user assisted sentiment analysis can be both costly and inefficient due to the Internet tariff and mobile devices´ inherent limitations. However, in public events and places, users can deploy a network over their connectivity enabled mobile devices to share sentiments at no monetary cost and analyze them to select a service. Yet, this network, feasible as a mobile ad hoc network, does not consider automatic incorporation of users´ sentiments for service selection. Therefore, we propose a service selection method that integrates users´ sentiments simultaneously with the traditional minimum hop count method to select a service in the mobile ad hoc network. These sentiments, classified by the Support Vector Machine with semantic pruning, have been quantified for integration. The novelty of the proposed selection method is further supported through simulation results.