CAARD - Context Aware App Recommendation and Delivery using Decision Support Systems

Objectives: To deliver the affinity of user(s) as a service to prospective developers and help them in providing context aware content. Methods: First, the list of apps installed on the mobile device is retrieved using a special module attached to the host application and is then tagged according to its genre. The module then runs a background service to record the active times of each application. All these information is synchronized with the database on the cloud periodically. The CAARD engine analyses the information in the database to predict the affinity of the user(s). Findings: Currently the users are classified based on their location and categorized using search history, browsing pattern which are not so efficient. They sometimes aim to deliver contents based on the web searches. This again does not necessarily mirror the requirements of the user as all the search terms may be trivial and need not always be specific to the user. Most if the previous systems aim to modify web based techniques for the mobile ecosystem which makes it less efficient. This directly reflects on the revenue of the developers or results in the fall of user base. The proposed system understands the user based on the applications that he frequently uses. This makes it even more user centric and helps the developers in delivering content that is specific to each of the users. Applications: One possible application of this system could be to display more relevant advertisements. This persuades the user to read and click the advertisement thus generating more revenue. Another implementation could be to recommend products that the user might be interested to buy.