1 of 11

Manish Parashar

Director, Scientific Computing & Imaging (SCI) Institute & Chair in Computational Science and Engineering

Presidential Professor, School of Computing

University of Utah

NSDF AHM, SLC, Utah, USA

October 11, 2022

What do we mean by democratizing data?

2 of 11

3 of 11

Event Horizon Telescope: Blackhole Image (Sagittarius A*)

Laser Interferometer Gravitational-Wave Observatory (LIGO) : Observation of Gravitational Waves

[LIGO image] https://www.ligo.caltech.edu/system/media_files/binaries/266/original/162571main_GPB_circling_earth3_516.jpg?1446243770

[EHT image] http://news.mit.edu/sites/mit.edu.newsoffice/files/styles/news_article_image_top_slideshow/public/images/2019/01-eso1907a_0.jpg?itok=LNUsoo9M

https://medium.com/starts-with-a-bang/this-is-why-the-event-horizon-telescope-still-doesnt-have-an-image-of-a-black-hole-192ea94f97b7

Search

data

Gather

data

Process

data

Observatory, Instrument, Repository

Scientific discovery

May 2022

2017 Nobel Prize

4 of 11

OSTP (Nelson) Memo (08/25/2022)

Ensuring Free, Immediate, and Equitable Access to Federally Funded Research

Remove the 12-month publication embargo on peer-reviewed articles resulting from Federally funded research;
Require that scientific data supporting the results reported in peer-reviewed articles are made available at the time of publication and to make plans for sharing other research data as well, ensuring privacy and security considerations are maintained;
Create new precedents for ensuring scientific and research integrity through transparency of Federally funded research with the use of persistent digital identifiers for all aspects of the research life-cycle (with extended deadlines);
Identify and address issues of inequality in Federally funded scholarly publishing; and,
Apply public access requirements to all Federal agencies with research and development expenditures, not just those originally subject to the 2013 guidance.

https://www.whitehouse.gov/wp-content/uploads/2022/08/08-2022-OSTP-Public-Access-Memo.pdf

(Update to 2013 Holdren Memo)

Increasing scale, heterogeneity, and richness of data

Data downloads and local processing no longer feasible

Largely disconnected from each other, from commercial/academic CI services

Integrating observatory data into science workflows is a growing challenge

New delivery modes for data and data-products essential

Multi-Messenger Astrophysics (MMA) : gravitational waves, particles, electromagnetic radiation

GW150914: blackhole merger: discovery made using the U.S.-based Laser Interferometer Gravitational-Wave Observatory (LIGO)
GW170817: neutron star merger: discovery made using LIGO; the Europe-based Virgo detector; and some 70 ground- and space-based observatories
IceCube-170922A: blazar TXS 0506+056 identified as source for extra-galactical neutrinos: discovery made using NSF’s IceCube Neutrino Observatory, NASA’s Fermi Gamma-ray Space Telescope, the Major Atmospheric Gamma Imaging Cherenkov Telescopes (MAGIC) in the Canary Islands, and about 20 ground- and space-based observatories

“On the Clock: Time lost is Science Lost!” – We have to identify the source as soon as possible, so that ground- and space-based observatories can begin observations asap. The latter is crucial to constrain and validate models and simulations. Critical for building predictive capabilities. The cosmic events of interest are explosive and the early stages are most challenging

Intellectual Merit: Multi-Messenger Astrophysics (MMA) is an exciting new field of science that combines traditional astronomy with the new capabilities to measure gravitational waves, high-energy neutrino particles and other messengers that originate from celestial objects. The promise of MMA can be realized only if sufficient cyberinfrastructure is available to rapidly handle, combine, and analyze the very large-scale distributed data from all the types of astronomical measurements.

5 of 11

Democratizing Science => Democratizing Data/Compute

Broad, fair, and equitable access to data/compute is essential to democratizing science

Significant barriers

Knowledge: Awareness, discovery, expertise, support
Technical: Allocation, access, on-ramps
Social: Awareness of the importance of access to CI, rewards structure

Open/equal access 🡺 democratized/equitable access

Data Discovery: Knowledge networks, intelligent data delivery
Data Access: Realtime, streaming, on-demand
Data Integration: Data integration & interoperability
Data-driven Science: Data processing, knowledge extraction

M. Parashar, "Democratizing Science Through Advanced Cyberinfrastructure" in Computer, vol. 55, no. 09, pp. 79-84, 2022. doi: 10.1109/MC.2022.3174928

https://www.computer.org/csdl/magazine/co/2022/09/09869577/1GeVwbmLj6U

Increasing scale, heterogeneity, and richness of data

Data downloads and local processing no longer feasible

Largely disconnected from each other, from commercial/academic CI services

Integrating observatory data into science workflows is a growing challenge

New delivery modes for data and data-products essential

Multi-Messenger Astrophysics (MMA) : gravitational waves, particles, electromagnetic radiation

GW150914: blackhole merger: discovery made using the U.S.-based Laser Interferometer Gravitational-Wave Observatory (LIGO)
GW170817: neutron star merger: discovery made using LIGO; the Europe-based Virgo detector; and some 70 ground- and space-based observatories
IceCube-170922A: blazar TXS 0506+056 identified as source for extra-galactical neutrinos: discovery made using NSF’s IceCube Neutrino Observatory, NASA’s Fermi Gamma-ray Space Telescope, the Major Atmospheric Gamma Imaging Cherenkov Telescopes (MAGIC) in the Canary Islands, and about 20 ground- and space-based observatories

“On the Clock: Time lost is Science Lost!” – We have to identify the source as soon as possible, so that ground- and space-based observatories can begin observations asap. The latter is crucial to constrain and validate models and simulations. Critical for building predictive capabilities. The cosmic events of interest are explosive and the early stages are most challenging

Intellectual Merit: Multi-Messenger Astrophysics (MMA) is an exciting new field of science that combines traditional astronomy with the new capabilities to measure gravitational waves, high-energy neutrino particles and other messengers that originate from celestial objects. The promise of MMA can be realized only if sufficient cyberinfrastructure is available to rapidly handle, combine, and analyze the very large-scale distributed data from all the types of astronomical measurements.

6 of 11

Democratizing Science => Democratizing Data/Compute

Broad, fair, and equitable access to data/compute is essential to democratizing science

Significant barriers

Knowledge: Awareness, discovery, expertise, support
Technical: Allocation, access, on-ramps
Social: Awareness of the importance of access to CI, rewards structure

Open/equal access 🡺 democratized/equitable access

Data Discovery: Knowledge networks, intelligent data delivery
Data Access: Realtime, streaming, on-demand
Data Integration: Data integration & interoperability
Data-driven Science: Data processing, knowledge extraction

M. Parashar, "Democratizing Science Through Advanced Cyberinfrastructure" in Computer, vol. 55, no. 09, pp. 79-84, 2022. doi: 10.1109/MC.2022.3174928

https://www.computer.org/csdl/magazine/co/2022/09/09869577/1GeVwbmLj6U

Increasing scale, heterogeneity, and richness of data

Data downloads and local processing no longer feasible

Largely disconnected from each other, from commercial/academic CI services

Integrating observatory data into science workflows is a growing challenge

New delivery modes for data and data-products essential

Multi-Messenger Astrophysics (MMA) : gravitational waves, particles, electromagnetic radiation

GW150914: blackhole merger: discovery made using the U.S.-based Laser Interferometer Gravitational-Wave Observatory (LIGO)
GW170817: neutron star merger: discovery made using LIGO; the Europe-based Virgo detector; and some 70 ground- and space-based observatories
IceCube-170922A: blazar TXS 0506+056 identified as source for extra-galactical neutrinos: discovery made using NSF’s IceCube Neutrino Observatory, NASA’s Fermi Gamma-ray Space Telescope, the Major Atmospheric Gamma Imaging Cherenkov Telescopes (MAGIC) in the Canary Islands, and about 20 ground- and space-based observatories

“On the Clock: Time lost is Science Lost!” – We have to identify the source as soon as possible, so that ground- and space-based observatories can begin observations asap. The latter is crucial to constrain and validate models and simulations. Critical for building predictive capabilities. The cosmic events of interest are explosive and the early stages are most challenging

Intellectual Merit: Multi-Messenger Astrophysics (MMA) is an exciting new field of science that combines traditional astronomy with the new capabilities to measure gravitational waves, high-energy neutrino particles and other messengers that originate from celestial objects. The promise of MMA can be realized only if sufficient cyberinfrastructure is available to rapidly handle, combine, and analyze the very large-scale distributed data from all the types of astronomical measurements.

7 of 11

Democratizing Science => Democratizing Data/Compute

Broad, fair, and equitable access to data/compute is essential to democratizing science

Significant barriers

Knowledge: Awareness, discovery, expertise, support
Technical: Allocation, access, on-ramps
Social: Awareness of the importance of access to CI, rewards structure

Open/equal access 🡺 democratized/equitable access

Data discovery
Data access
Data processing, integration
Data-driven science

M. Parashar, "Democratizing Science Through Advanced Cyberinfrastructure" in Computer, vol. 55, no. 09, pp. 79-84, 2022. doi: 10.1109/MC.2022.3174928

https://www.computer.org/csdl/magazine/co/2022/09/09869577/1GeVwbmLj6U

Increasing scale, heterogeneity, and richness of data

Data downloads and local processing no longer feasible

Largely disconnected from each other, from commercial/academic CI services

Integrating observatory data into science workflows is a growing challenge

New delivery modes for data and data-products essential

Multi-Messenger Astrophysics (MMA) : gravitational waves, particles, electromagnetic radiation

GW150914: blackhole merger: discovery made using the U.S.-based Laser Interferometer Gravitational-Wave Observatory (LIGO)
GW170817: neutron star merger: discovery made using LIGO; the Europe-based Virgo detector; and some 70 ground- and space-based observatories
IceCube-170922A: blazar TXS 0506+056 identified as source for extra-galactical neutrinos: discovery made using NSF’s IceCube Neutrino Observatory, NASA’s Fermi Gamma-ray Space Telescope, the Major Atmospheric Gamma Imaging Cherenkov Telescopes (MAGIC) in the Canary Islands, and about 20 ground- and space-based observatories

“On the Clock: Time lost is Science Lost!” – We have to identify the source as soon as possible, so that ground- and space-based observatories can begin observations asap. The latter is crucial to constrain and validate models and simulations. Critical for building predictive capabilities. The cosmic events of interest are explosive and the early stages are most challenging

Intellectual Merit: Multi-Messenger Astrophysics (MMA) is an exciting new field of science that combines traditional astronomy with the new capabilities to measure gravitational waves, high-energy neutrino particles and other messengers that originate from celestial objects. The promise of MMA can be realized only if sufficient cyberinfrastructure is available to rapidly handle, combine, and analyze the very large-scale distributed data from all the types of astronomical measurements.

8 of 11

The Virtual Data Collaboratory

Goals/features:

A dedicated high-speed network, compute and storage resources federated over the participating sites
Data discovery and access: a set of Data Services including indexing, cataloging, sharing and metadata management according to FAIR principles
An Internet-scale execution platform, allowing to distribute complex distributed analytics close to the data source
A network of high-performance DTNs equipped with a middleware for implementing smart data discovery, delivery, and management strategies.
Train the next generation of scientists with deep disciplinary expertise in leveraging data, cyberinfrastructure, and tools to address research problems.

M. Parashar et al., "The Virtual Data Collaboratory: A Regional Cyberinfrastructure for Collaborative Data-Driven Research," in Computing in Science & Engineering, vol. 22, no. 3, pp. 79-92, 1 May-June 2020, doi: 10.1109/MCSE.2019.2908850.

https://www.computer.org/csdl/magazine/cs/2020/03/08686134/1hXdyNXrgt2

9 of 11

The Virtual Data Collaboratory (See Ivan Rodero’s Talk)

M. Parashar et al., "The Virtual Data Collaboratory: A Regional Cyberinfrastructure for Collaborative Data-Driven Research," in Computing in Science & Engineering, vol. 22, no. 3, pp. 79-92, 1 May-June 2020, doi: 10.1109/MCSE.2019.2908850.

https://www.computer.org/csdl/magazine/cs/2020/03/08686134/1hXdyNXrgt2

10 of 11

Towards Intelligent Data Discovery and Delivery

Goal: Explore how an understanding of the data and their usage, can be used to democratize access to LF data/products

Data Delivery

Analyse LF data and data access
In-network data caching, prefetching

Data Discovery:

Data/data product recommendations

Builds on edge, in-network (DTN)

capabilities

Y. Qin, I. Rodero and M. Parashar, "Toward Democratizing Access to Facilities Data: A Framework for Intelligent Data Discovery and Delivery" in Computing in Science & Engineering, vol. 24, no. 03, pp. 52-60, 2022.�doi: 10.1109/MCSE.2022.3179408

Introduce the research purposes
Three studies

Analyzing data access patterns lay out the foundation for the data delivery and discovery service
Intelligent data delivery service: leveraging pre-fetching and caching to accelerate user data accessing performance
Intelligent data discovery service: using the user data access affinities and knowledge graph techniques to build up the CKAT recommendation model

Overview of the framework
It builds on top of DMZ network

High-throughput network
Optimized for transferring large-scale scientific data
DTN:

Accessing point to DMZ network
Have computing power and large storage

Framework components

Pattern learning module analyzes the user access pattern, and pass the knowledge to data discovery and delivery service

Data discovery service

Recommendation system
Push recommended data/data products to users

Data delivery service

Pre-fetching data
Cache data
Accelerate data access performance

1 of 11

2 of 11

3 of 11

4 of 11

5 of 11

6 of 11

7 of 11

8 of 11

9 of 11

10 of 11

11 of 11