CAREER: Graph Information System: Deciphering Complex Networks

图形和网络无处不在，编码从化学键到社会互动的复杂关系。 在这些网络中隐藏着生物学、商业和社会学中许多重要问题的答案。 为了分析复杂的网络，用户必须掌握复杂的计算和编程技能。 本项目旨在通过开发一个通用的图信息系统来改变现有技术的现状，以满足复杂网络搜索和挖掘的需求。 现实生活中的网络是复杂的，不仅具有拓扑结构，而且还包含异构的内容和属性与节点和边。结构和内容的混合提出了两个挑战，需要新的解决方案来实现更智能、更快速的图形分析。 首先，出现了新类型的图搜索和挖掘操作，例如图聚合、图关联和图模式挖掘。 其次，当图变得复杂和庞大时，大多数现有的图挖掘算法不能很好地扩展。 本项目针对这些挑战，并进行了全面的研究，一般图形信息系统。 该系统包括三个主要部分：复杂图搜索，图模式挖掘和图索引。 它涵盖了社会，生物和信息网络中新兴的结构查询，新的图挖掘操作符，如图摘要和关联，以及创新的索引方法，例如，这项研究通过学生指导和课程开发与教育紧密结合。该项目产生的出版物、软件和课程材料在项目网站上传播：http://www.cs.ucsb.edu/~xyan/gis.html。

Graphs and networks are ubiquitous, encoding complex relationships ranging from chemical bonds to social interactions. Hidden in thesenetworks are the answers to many important questions in biology,business, and sociology. In order to analyze complex networks, users have to master sophisticated computing and programming skills. It indeed becomes a pain point for many scientists and engineers.This project is to change the state of the art by developing a general graph information system, which is able to address the needs of searching and mining complex networks. Real-life networks are complex, not only having topological structures, but also containing heterogeneous contents and attributes associated with nodes and edges. The mixture of structures and contents raises two challenges that require new solutions for smarter and faster graph analysis. First, new types of graph search and mining operations, such as graph aggregation, graph association, and graph pattern mining, are emerging. Second, when graphs become complex and large, most of existing graph mining algorithms cannot scale well. This project addresses these challenges and performs a comprehensive study of a general graph information system. The system includes three major components: complex graph search, graph pattern mining, and graph indexing. It covers emerging structure queries in social, biological, and information networks, new graph mining operators such as graph summarization and association, and innovative indexingmethodologies, e.g., differential graph index.This research is tightly integrated with education through student mentoring and curriculum development. Publications, software and course materials resulted from this project are disseminated on the project website: http://www.cs.ucsb.edu/~xyan/gis.html.