大数据的智能处理迫切需要语义技术的支持，然而现有的大数据技术平台都相对缺乏数据语义的处理能力。本项目主要目标是研究面向大数据的语义计算框架，并推动其在生物医学大数据等领域的应用。具体主要围绕三个科学问题:a)怎样平衡数据的语义表达和弹性可扩展的数据处理能力？b)怎样实现跨边界跨尺度的大数据语义挖掘与知识发现？c)怎样实现更加智能的大数据自然语义交互。并拟开展四个方面的研究，包括：a)研究面向大数据特征的语义表达框架；b)研究跨边界、跨尺度的大数据语义发现方法；c)研究弹性可扩展的语义大图处理引擎及高实时的语义交互技术；d)结合生物医学和健康领域的大数据处理开展应用创新研究。同时，形成三个方面的关键技术创新：a)轻语义表达模型；b)跨尺度语义挖掘；c)可演化语义大图存储与实时语义交互。随着图模型在大数据应用中日益重要,本项目的研究有助于提升在大数据核心技术领域的竞争力。

More advanced and intelligent big data processing needs semantic technologies urgently. However, current big data platforms lack the capability of processing data semantics. The principal goal of the project is to develop a semantic computing framework geared toward typical characteristics of big data applications, and foster its application in particular domains such as biomedical informatics. Specifically, we consider three main problems including: a) How to balance the expressiveness of data semantics and scalability of data parallelism? b) How to discover deep semantic relations from multi-scale big data? c) How to enable real time semantic interaction with big data? The research consists of four sub-topics including: a) Semantic model for big data; b) Cross-boundaries and multi-scale semantic discovery from big data; c) Elastic and scalable engine for big semantic graph storage and query interactions; d) Applications in biomedical knowledge discovery and biomedical semantic network processing. The key innovations of the project lie in three aspects: a) A light-weight semantic model for big data; b) Multi-scale semantic discovery for big data; c) Evolvable storage and real time query engine for big semantic graph.