对于缺乏安卓应用开发经验的开发者而言，正确地调用API方法来实现应用的功能特征往往会花费大量时间。代码推荐与完成技术能根据已编写的代码向开发者推荐已有应用的相关代码片段，并进一步填充后续代码。然而，现有的基于模式的方法并未结合安卓应用本身所具有的覆盖特定主题、不提供源代码和碎片化的特性。因此，本项目将针对以上特性，强调语义信息在推荐与完成技术中的作用，探索更加切合安卓应用开发的代码推荐与完成技术。本项目的研究工作围绕安卓应用中普遍存在的逻辑模式进行展开，逻辑模式是实现某一特定安卓功能特征（主题）的、跨越不同类的API调用序列。本项目的研究目标是显式地描述和管理安卓应用的逻辑模式，研究有效地从大量的安卓应用中挖掘并整理逻辑模式的方法，同时研究支持安卓代码推荐与完成的代码上下文分析、基于主题语义的模式推荐与代码模板实例化的技术，从而为面向安卓应用开发的代码推荐与完成提供理论依据和技术保证。

It always takes a lot of time for developers lacking of experience in Android development to invoke the API methods correctly in order to implement a functional feature. The techniques of code recommendation and completion can help to suggest related code snippets and further to fill the subsequent codes according to the existing code. However, available pattern-based approaches have not incorporated the characteristics of covering specific themes, not providing source code and fragmentation in Android applications. To this end, this project aims to emphasize the role of semantic information in code recommendation and completion, and to explore the recommendation and completion techniques that are more suitable for Android application development. The research work is carried out around the logic pattern commonly existing in Android applications which is defined as a set of API invocation sequences across different classes that are implemented to achieve a particular Android feature (theme). The research target of this project is to explicitly describe and manage the logic patterns of Android applications, to investigate the approach of effectively mining and re-organizing the patterns from a large number of applications, and to work on the methods of code context analysis, theme-based pattern recommendation and code template instantiation that support Android development. The research achievement is anticipated to provide a theoretical basis and technical guarantee for pattern driven code recommendation and completion in Android application development.