Building efficient mobile systems and applications for supporting information exchange in resource limited settings

Abstract

Enabling support for information exchange has a great impact on the development of different sections of the society. Mobile phones using telecommunication services have enabled such support in several scenarios including healthcare, agriculture, and physical safety. In many scenarios, mobile phones are also the preferred technological medium like reaching out to police or hospital for assistance in emergency situations. In several geographies such as rural areas of developing regions, mobile phones are the only possible solution due to several human factors (illiteracy, poverty, etc.) and poor infrastructure support for other technological mediums. This motivated us to study different systems that support information exchange using mobile phones or telecommunication services.

In this thesis, we focus on building efficient mobile-based information sharing technologies for their benefits to different sections of society including developing regions. We have study three basic forms of the telephonic system that are currently used for information exchange and other telephonic systems can be realized by combining these three systems. First, we study helpline systems that are mostly used as an emergency responder and customer care systems. We developed a statistical method to evaluate and design temporal metaphors that can help to reduce the number of unanswered calls on a helpline system. We evaluated our solution using a series of real-world experiments conducted at Government Helpline. Second, we study automated voice applications that are widely used in ticketing services of airline and railways. Our proposed adaptive voice response system provides a way to reduce navigation time so that desired information can be accessed quickly on menu based voice application. We have also presented the real word evaluation of our adaptive system. Third, we study Smartphone applications that are also becoming popular in several areas including remote health monitoring and sensing. Our proposed method provides a solution for smartphone based deployments in regions where data connectivity is a substantial hurdle to such implementations. Towards this, we have build a system that can reduce the dependency on data connectivity by enabling multimodal communication in the telephonic system. We have tested our system in a rural area over a period of nine months. We have also developed and evaluated method to send data through voice when data channels are not available with the cellular network. A prototype for android platform is also developed along the same line to send SOS signal with GPS data from smartphone application when data connectivity is not available.