CAREER: Cyber-Knowledge Infrastructure for Geospatial Data

Project: Research project

Description

Overview
This career development plan seeks to build the foundation of the PI's long-term career focusing on the integrated research and education of C-KING, a cyber-knowledge infrastructure, that enhances access, search and reasoning of distributed big geospatial data. Since the amount of geospatial data keeps increasing at a unprecedented speed and is widely scattered in the cyberspace, it is critical for the next-generation cyberinfrastructure to provide ready access to the best available information and to provide spatiotemporal reasoning ability on top of these data to foster knowledge synthesis. This project will lead to (1) an in-depth understanding of the cyber-distribution pattern of geospatial resources on the ever-expanding Web; (2) a new way of effective geospatial data search that serves as a core component in future cyberinfrastructure, and (3) a C-KING, which extends traditional cyberinfrastructure for knowledge management and deploys spatiotemporal reasoning ability to intelligently answer questions related to the Earth's surface.
Intellectual Merit: The intellectual merit of this proposal is a body of knowledge regarding the theoretical and practical aspects of a C-KING that can answer complex spatiotemporal questions utilizing the best available geospatial data. One of the most important barriers in performing science lies in that current data search tools are mostly built upon standard techniques from computer science. These methods ignore the inherent properties (location awareness, structural and semantic heterogeneity) of geospatial data, leading to high false negative and false positive search results. This proposal describes two innovative approaches: large-scale web mining as well as a geospatial semantic search to address this long-standing problem. The large-scale web mining substantially increases the accessibility of available geospatial data resources. A hybrid geospatial semantic search significantly improves the discoverability of these collected resources. On top of this data search component, this proposal develops a spatiotemporal reasoning ability that integrates gazetteer, temporal reasoner, workflow engine, and a suite of spatial analysis routines for intelligent question answering related to space and time. This research has significant merits as follows. First, it designs effective algorithms and models within a search tool to make geospatial data more discoverable and accessible. Second, it moves scientific workflows which require extensive GIS (Geographic Information Science) skills into a cyberinfrastructure environment. As such, within C-KING, different data, analysis tools and workflows can be easily accessed and reused, spatiotemporal questions can be timely answered, and the scientific results can be reproduced and validated on-the-fly. The proposed work is rooted in the PI's vast experience and knowledge in GIScience, cyberinfrastructure, spatial information retrieval, semantic interoperability, metadata and distributed computing.
Broader Impacts: The demand of geospatial data from many fields, including geography, earth science, geoscience, polar science, atmospheric science etc, has been driving the need for increased data access and productivity. This need is especially urgent as big data becomes dominant in the science landscapes. The C-KING project, which supports large-scale automated collection, smart search and complex spatiotemporal analysis of geospatial data, fits right into this vision. If successful, the research outcomes will pave the road for developing more intelligent cyberinfrastructure to support data and knowledge discovery in a variety of data-intensive applications. The PI's prior work on this theme has been well recognized by the open geospatial consortium, research institutions and industry. Some integration work is already going on. The PI plans to take advantage of this established collaboration network to (1) broaden the use of the developed technique; and (2) collect feedback from peers to continuously improve the proposed work. All software modules developed through the C-KING project will be made open-source.
The PI was recently selected as a NSF-sponsored CyberGIS fellow and she will actively contribute to the community-driven geospatial cyberinfrastructure education and outreach, including: (1) training the next generation cyberinfrastructure researcher and engineer through curriculum development and involvement of undergraduates in research, (2) increasing the participation of underrepresented groups in C-KING research, and (3) exciting K-12 students with cyberinfrastructure supported geospatial sciences.
StatusActive
Effective start/end date5/1/154/30/20

Funding

  • National Science Foundation (NSF): $449,859.00

Fingerprint

infrastructure
workflow
science
semantics
ability
career
resources
data access
curriculum development
available information
virtual reality
knowledge management
information retrieval
information science
computer science
knowledge
engineer
education
data analysis
productivity