MSleepWatcher: Why didn't i sleep well?

It has been medically proven that one's sleep quality directly affects his or her personal health, social behavior, and work effectiveness. Understanding sleep quality, hence, is an important topic and has been attracting a vast amount of research efforts. In addition, with the advent and ever-increasing number of mobile and wearable devices, many attempts have been made towards monitoring one's sleep using these ubiquitous devices. However, there were none on exploring the complex relationship between sleep quality and sleeping environment. In this paper, we propose mSleepWatcher as an on-going attempt answer the question of "Why didn't I sleep well?". By mining the environmental sensing information collected by built-in sensors of off-the-shelf mobile and wearable devices in combination with the sleep quality sensing information, a set of causality analysis techniques is adopted and applied to exploit the existence of temporal dependencies between the environment during sleep and sleep quality. Resulted from mSleepWatcher system, latent relationships between the environment and sleep quality can be inferred which are then used to provide recommendation to users to suggest users adjusting their sleep environment for a better sleep. The proposed system is the first attempt to bring a fresh picture of sleep study associated with different scenarios of environmental variations. Derived from our preliminary work, both strength and limitations for developing the complete system in mobile devices are discussed in detail.