A recommendation system for emergency mobile applications using context attributes: REMAC
Alireza Ahmadi, Debjyoti Mukherjee, Guenther Ruhe · 2019
The extensive use of mobile devices had led to tremendous growth in not only the usage of different apps but also their capability to help people in moments of crisis. There are different emergency mobile apps published in the app markets; these apps can be of enormous assistance to victims as they can provide valuable information and guidance at the opportune moments. However, app store reviews, ratings, and relevant studies have revealed that users are often averse to using these apps or their different features. This draws our attention to the need for recognizing essential features and including them in the emergency apps to increase their usability. Our proposed recommendation system called REMAC combines different machine learning techniques to analyze the context characteristics of different organizations and suggest unique features that can be included in their emergency apps. REMAC is built by analyzing 24 potential context attributes of 1909 universities spread across North America. This research also includes a systematic attribute selection process that enables us to reach a local optimum for the given dataset. This tool carefully dissects the context attributes of each university and suggests top features that should be included in its emergency app. It leverages the data (other apps and features) provided by the app markets to suggest essential features.