社会移动云环境下的APP自主协同及个性化混搭机制研究

Focus on the new challenge that user requirements tends to complex and is hard to deal with a single fragmentation mobile application (APP) in the Social Media, Mobile Internet and Cloud Computing (SMC) environment, this project will study the key issues of adaptive collaboration and chain-reactively invoking autonomous heterogeneous APPs to satisfy user complex requirements through mobile devices. By collecting and analyzing APP data and digital footprints in SMC environment, the semantic relations among APPs are mined, and a complex network for APPs ecosystem is are modeled and designed. And then utilizing Social Network Analysis (SNA), Data Mining and Machine Learning, we will study the characteristics, the dynamic adaptive collaboration principle, and the evolution mechanism of APPs ecosystem. Subsequently, we will study the recommendations for personalized APP Mashup based on dynamic collaboration and ecological characteristics, and design the formal interface specification, adaptive message bus, chain-reactively execution engine which will support the adaptive information sharing and interoperability between heterogeneous APPs. Then there is a framework which is built on Crowdsourcing and designed to facilitate end-user without good training in programming developing, sharing, reusing and personalizing of Mashup instance. In short, this project would exploit an approach that various heterogeneous mobile APPs adaptively collaborate and cooperate each other, automatically execute one by one without user intervention to satisfy user complex requirements through mobile devices anytime and anywhere. Thereby, as far as we know, it will of great significance for smoothing user interaction, improving user experience, creating value-added by end-user, promoting the healthy progress and prosperity of the whole mobile APP ecosystem.

针对社会移动云环境下单一移动应用(APP)难以满足用户复杂需求现状，本项目研究面向APP自主协同连锁式响应执行的关键问题。通过对APP和用户数字足迹的收集及分析，挖掘异构自治APP间的语义关联、构建APP生态网络系统模型。基于网络分析、数据挖掘和机器学习研究APP网络的生态系统特性、动态协同机理及演化策略，研究基于生态网络和动态协同机理的个性化协同方案推荐方法，研究设计支持异构APP自适应信息共享与互操作的接口规范、自适应消息总线、连锁反应式触发执行引擎，研究基于众包的协同实例构建和分享框架，从而提供经由异构自治APP自主协同、静默式连锁执行的“一揽子”解决方案来满足用户随时随地通过智能终端处理复杂任务的需求。减少用户显式交互、改善用户体验，使每个用户都成为混搭方案的提供者，对APP价值增值及整个移动应用生态系统的繁荣有重要意义。

社会移动云环境下APP应用通过用户手工驱动的方式来响应用户的复杂需求，缺乏自主性、个性化和灵活性。针对该问题，本项目在解决孤立异构APP的信息共享和传输基础上，通过对APP数据、数字脚印等的挖掘分析发现APP的协同关联，实现“异构自治APP连锁反应式动态协同响应用户复杂需求”。本项目主要从以下3个方面展开深入研究：.1）移动应用生态系统建模和语义关系挖掘：本项目提出了一种基于WordNet的多层异构复杂应用程序网络模型，为从复杂网络角度研究异构APP动态协同问题奠定基础。同时，为了有效地组织和利用多源异构数据，挖掘APP间的语义交互和关联关系，本项目还提出了本体匹配框架和一种增量评估和纠正web服务的语义标注的技术，分别用于精确识别不同粒度级别的本体匹配和提高语义标注的质量。.2）异构APP类生态特性及协同演进机理研究：本项目在APP语义网络的基础上，通过复杂网络理论和分析方法，刻画异构APP语义网络的类生态系统特性，并提出了基于网络嵌入的广义社区检测方法和节点粒度增量社区检测方法。此外，基于分布式知识的服务生态系统演化模型，研究基于APP的动态协同和演化机理，为面向解决方案的个性化动态协同混搭“APP工作流”的推荐与构建奠定基础和提供指导。.3）APP协同混搭方案推荐及互操作：本项目综合考虑APP的社会化交互特性、协作模式、生态网络结构及协同机理等多种因素，将传统的推荐单一移动应用优化为推荐由一系列“相关APP”协同&混搭而成的解决方案。以APP生态网络中信息流动研究为基础，优化Web服务的互操作性，并提出了一种基于微服务架构的跨界服务融合方法，优化异构APP间自适应信息共享与互操作，最后，提出一种基于APP无缝集成的应用流自动协同框架和基于细粒度APP的客户端微服务自动协同框架，实现APP应用连锁反应式静默执行机制。