Dr. Line Pouchard

Projects

Projects

Provenance-enabled sample measurements for multi-modal analysis

New generation detectors produce massive amounts of data that require novel infrastructure and advanced computing discovery and analysis techniques to address the challenges of designing and validating new material structures. In this project, we provide building blocks compatible with BlueSky, the NSLSII Data Acquisition system. We provide heterogenous searches across beamlines, sample metadata, and search/match for pair-distribution functions over the Crystallography Open Database. (BNL LDRD PPSL 2018-005)

DOE Exascale Computing Program CODAR: Online Data Analysis and Reduction (ECP 17-SC-20-SC)

The CODAR co-design center's goal is to co-optimizes simulation, analysis, reduction for performance and information output. This is new to the traditional simulation-output-analyze approach. In this project we use Provenance to extract information enabling detailed introspection of computational workflows at runtime and extend its use to performance analysis (2017-20, PI: Ian Foster).

CIF21 DIBBs: Building a Modular Cyber-Platform for Systematic Collection, Curation, and Preservation of Large Engineering and Science Data - A Pilot Demonstration Project. (NSF, 2014-2016)

DataCenterHub provides a simple, standardized yet flexible platform to preserve and share scientific and engineering data produced by measurements and observations of structures undergoing stimuli in civil engineering. We are improving scholarly communication in DataCenterHub by developing a metadata corpus and building co-author networks on issues of interest to the community (COMPLETED).

ESIP Federation Semantic Portal

The ESIP Semantic Portal provides a repository for Earth Sciences ontologies developed in OWL. In collaboration with South Carolina University, the portal deploys the NCBO Bioportal virtual appliance (COMPLETED).

DataONE: the Data Observation Network for Earth. Co-founder and interim leader of the Data Integration and Semantics Working group.

DataONE is the foundation of new innovative environmental science through a distributed framework and sustainable cyberinfrastructure that meets the needs of science and society for open, persistent, robust, and secure access to well-described and easily discovered Earth observational data. I improved data discovery in DataONE by 1)providing linkages between an ontology repository and the Mercury search system, 2)using ontologies to expand queries, and 3)performing text mining to enrich DataONE metadata sets. Sponsor:NSF (COMPLETED)

Providing Rich Metadata Annotations for Earth Science Data. Remote Data Analysis and Visualization Center

This effort seeked to enrich keywords in metadata documents by pairing each document with another using various cosine similarity methods and cross-populating keywords between members of the pair. Sponsor:NSF (COMPLETED)

Enhancing Climate Impact Integrated Assessment for Water through Climate Informatics.

Climate Informatics is the science of information applied to the study of climate change, its consequences and its impacts. This project focuses on the integrated transformation of data and processes with an end-to-end, systemic approach, an emphasis on traceable, quality-controlled information, and scenario-based analyses using both observation and modeling data. Sponsor: ORNL LDRD (COMPLETED).

A Framework for the Collection and Analysis of Large Datasets

In this project, DCAT enables the collection, storage, analysis, and display of sensor data to detect network and processor failure in the next generation of HPC. RSS feeds are used to stream data from heterogeneous sources. DCAT is currently undergoing parallelization to run on a large cluster of Symmetric Multiprocessing (SMP) machines. Sponsor: DOD, Extreme Scale Systems Center, ORNL (COMPLETED).

Collection and Analysis Tools for the Intelligence Community (ORCAT I and II)

ORCAT I and II are lightweight, desktop tools with a smooth, easy-to-use interface. ORCAT I integrates functionality for building searchable, personal, metadata-rich collections of Web content with the ability to perform sophisticated queries. ORCAT II resolves open-ended queries and highlights hidden relationships using RDF. Both were developed to facilitate the use of OSINT. Acknowledgements: DOE Office of Science, Science Undergraduate Laboratory Internship (COMPLETED).

Group Violent Intent Modeling (GVIM)

The ORNL team participated in the R & D of the GVIM Intelligence Framework under the leadership of PNNL. At ORNL, we developed social science ontologies, provided all intellectual content for a Social Science Guidance Roadmap, and played an active role in the development of a decision-assistance tool for the Intelligence Community. Sponsor: DHS (COMPLETED).

Data model for end-to-end simulations with Leadership Class Computing

Simulations, data and experts seamlessly participate as part of an end-to-end interactive and adaptive discovery process, similar to conventional experimental laboratories. We prototyped the CAFÉ common data model for storing and organizing provenance information and hardware profiles of a given run. Cafe included metadata for Combustion (S3D), Astrophysics (Chimera), Fusion GTC/XGC, and Environment) models. Sponsor: DOE Office of Science (COMPLETED).

The Virtual Autopsy

The Virtual Autopsy project collected image data from combat fatalities to research causes of trauma and better design protection systems for soldiers. We developed GWAVA, an Information Retrieval Web application for the Virtual Autopsy. Sponsor: DARPA. Acknowledgements: DOE Office of Science, Science Undergraduate Laboratory Internship (COMPLETED).

A Knowledge-Based Middleware and Visualization Framework for the Virtual Soldier

The Virtual Soldier Project investigated methods to represent and predict physiological, electro-mechanical, and anatomical properties of individual medical data on an electronic "dog tag" for assisting with care on the battlefield. The ORNL team addressed complex aspects of data, information, and knowledge integration across VSP teams and modules as required for the modeling and visualization of a disturbance of anatomical and physiological properties. Sponsor: DARPA (COMPLETED).

Sharing a World of Data: Scaling the Earth System Grid to Petascale

ESG is a project of the U.S. Department of Energy Scientific Discovery through Advanced Computing program. ESG addresses the formidable challenges associated with enabling analysis of and knowledge development from global Earth System models. ESG is a virtual collaborative environment that links distributed centers, users, models, and data. My contribution was the research and design of prototype ontologies and active participation in the development of the ESG metadata schema (COMPLETED).

Process Specification Language

PSL is an ontology and a specification language written in the Knowledge Interchange Format (KIF) originally designed to express manufacturing and design processes. PSL became a standard from the International Standard Organization (ISO 18629), in the Industrial Automation Systems and Integration, Industrial Data (SC4). Sponsor: the National Institute of Standards and Technology (NIST). (COMPLETED).