A Platform for Searching UI

Android applications are becoming increasingly popular in digital home nowadays. Building high quality application is still one of biggest changes for Android developer community. The search service in current Android application market can’t meet the needs of developers when trying to find the user interface design in other applications. In this paper, we introduce a platform to query the runtime user interfaces of Android applications, which can help developers to reuse the design components to improve their applications. Based on the description information of 8,456 Android applications in Android market and the information of runtime 313,820 UI elements, we implement the platform providing search service for developers to find specified runtime user interfaces in our Android software repository. Our experiment for the quality of search result shows that 76.7% of top-10 search result contains the relevant UI components that are valuable for users

For software developers, they usually look at some similar applications before designing an app. However, for the UI developers of Android app, it is hard for them to find reference user interfaces. For example, when developerswant to find some interfaces related to a specified topic“sign in”, they may search the apps in android market, install the related app and find the Sign-In interface manually. It’s very inefficient because there are a huge number of applications. First, it is hard to find the apps containing user interface for providing a specific service. Existing Android app markets provide app searching by matching keywords with apps’ description. It is hard to search a specific service especially for service which is not related to the core functionality of an app. Second, finding the user interface is time-consuming. The developer should download and install the application, and then explore the app mutually. In this paper, we propose an approach to search user interfaces(UI) in order to help developers when designingand maintaining their applications. We develop an platform to download and run the android apps and collect its runtime information by synchronization computers. For each app, our approach collects its description information in Android .CodeShoppy

market and its runtime information to save in the database. Based on the data collected, we develop a novel reverted index for user interface search. Our search engine provides the screen shots of runtime UI so that developers can learn from the design information and develop high-quality Android mobile efficiently. The search engine is implemented base on the information of 8,456 Android apps. In the experiment, we evaluate the quality of search result. The experiment shows that 76.7% of top-10 search result contains the relevant UI components that are valuable for users.The rest of this paper is organized as follows. In Section2, we introduce how to collect android application data. In Section 3, we illustrate the implementation the search engine.In Session 4 we present the experimental result. We review the related work in session 5. Finally, the paper is concluded in Section 6.