BigData Analysis of Stack Overflow for Energy Consumption of Android Framework

With the recent increase in the sales of Android-based smart-phones, the demand for android development from the past years has improved tremendously. The recent debacle of Samsung Galaxy Note 7 along with some other incidents have highlighted the need for better energy management in smart-phones. One source of the reported battery issues was the energy-inefficient APIs used by the Android developers. To improve reliability and better power utilization this needs to be addressed by the developers. This paper reports our analysis of this problem. In our study, we have used NLTK techniques, for analyzing energy-related posts on Stack Overflow (SO). Energy inefficient APIs have been discussed frequently by the developers on SO. We have analyzed over four million posts from SO from which thousands of posts on the topic of energy inefficient APIs are scrutinized in detail. Our study identifies the issues that are frequently related to LocationListener API, LocationManager API, Broadcastreceiver API in term of GPS location, and AlarmManager for location updation.