Universiti Teknologi Malaysia Institutional Repository

An optimized context-aware mobile computing model to filter inappropriate incoming calls in smartphone

Davoudi, Vahid (2020) An optimized context-aware mobile computing model to filter inappropriate incoming calls in smartphone. PhD thesis, Universiti Teknologi Malaysia, Razak Faculty of Technology & Informatics.


Official URL: http://dms.library.utm.my:8080/vital/access/manage...


Requests for communication via mobile devices can be disruptive to the receiver in certain social situation. For example, unsuitable incoming calls may put the receiver in a dangerous condition, as in the case of receiving calls while driving. Therefore, designers of mobile computing interfaces require plans for minimizing annoying calls. To reduce the frequency of these calls, one promising approach is to provide an intelligent and accurate system, based on context awareness with cues of a callee's context allowing informed decisions of when to answer a call. The processing capabilities and advantages of mobile devices equipped with portable sensors provide the basis for new context-awareness services and applications. However, contextawareness mobile computing systems are needed to manage the difficulty of multiple sources of context that affects the accuracy of the systems, and the challenge of energy hungry GPS sensor that affects the battery consumption of mobile phone. Hence, reducing the cost of GPS sensor and increasing the accuracy of current contextawareness call filtering systems are two main motivations of this study. Therefore, this study proposes a new localization mechanism named Improved Battery Life in Context Awareness System (IBCS) to deal with the energy-hungry GPS sensor and optimize the battery consumption of GPS sensor in smartphone for more than four hours. Finally, this study investigates the context-awareness models in smartphone and develops an alternative intelligent model structure to improve the accuracy rate. Hence, a new optimized context-awareness mobile computing model named Optimized Context Filtering (OCF) is developed to filter unsuitable incoming calls based on context information of call receiver. In this regard, a new extended Naive Bayesian classifier was proposed based on the Naive Bayesian classifier by combining the incremental learning strategy with appropriate weight on the new training data. This new classifier is utilized as an inference engine to the proposed model to increase its accuracy rate. The results indicated that 7% improvement was seen in the accuracy rate of the proposed extended naive Bayesian classifier. On the other hand, the proposed model result showed that the OCF model improved the accuracy rate by 14%. These results indicated that the proposed model is a hopeful approach to provide an intelligent call filtering system based on context information for smartphones.

Item Type:Thesis (PhD)
Uncontrolled Keywords:Improved Battery Life in Context Awareness System (IBCS), context-awareness, Optimized Context Filtering (OCF)
Subjects:T Technology > TK Electrical engineering. Electronics Nuclear engineering
Divisions:Razak School of Engineering and Advanced Technology
ID Code:97940
Deposited By: Yanti Mohd Shah
Deposited On:07 Nov 2022 18:57
Last Modified:07 Nov 2022 18:57

Repository Staff Only: item control page