Members Can Post Anonymously On This Site
-
Similar Topics
-
By NASA
6 min read
Smarter Searching: NASA AI Makes Science Data Easier to Find
Image snapshot taken from NASA Worldview of NASA’s Global Precipitation Measurement (GPM) mission on March 15, 2025 showing heavy rain across the southeastern U.S. with an overlay of the GCMD Keyword Recommender for Earth Science, Atmosphere, Precipitation, Droplet Size. NASA Worldview Imagine shopping for a new pair of running shoes online. If each seller described them differently—one calling them “sneakers,” another “trainers,” and someone else “footwear for exercise”—you’d quickly feel lost in a sea of mismatched terminology. Fortunately, most online stores use standardized categories and filters, so you can click through a simple path: Women’s > Shoes > Running Shoes—and quickly find what you need.
Now, scale that problem to scientific research. Instead of sneakers, think “aerosol optical depth” or “sea surface temperature.” Instead of a handful of retailers, it is thousands of researchers, instruments, and data providers. Without a common language for describing data, finding relevant Earth science datasets would be like trying to locate a needle in a haystack, blindfolded.
That’s why NASA created the Global Change Master Directory (GCMD), a standardized vocabulary that helps scientists tag their datasets in a consistent and searchable way. But as science evolves, so does the challenge of keeping metadata organized and discoverable.
To meet that challenge, NASA’s Office of Data Science and Informatics (ODSI) at the agency’s Marshall Space Flight Center (MSFC) in Huntsville, Alabama, developed the GCMD Keyword Recommender (GKR): a smart tool designed to help data providers and curators assign the right keywords, automatically.
Smarter Tagging, Accelerated Discovery
The upgraded GKR model isn’t just a technical improvement; it’s a leap forward in how we organize and access scientific knowledge. By automatically recommending precise, standardized keywords, the model reduces the burden on human curators while ensuring metadata quality remains high. This makes it easier for researchers, students, and the public to find exactly the datasets they need.
It also sets the stage for broader applications. The techniques used in GKR, like applying focal loss to rare-label classification problems and adapting pre-trained transformers to specialized domains, can benefit fields well beyond Earth science.
Metadata Matchmaker
The newly upgraded GKR model tackles a massive challenge in information science known as extreme multi-label classification. That’s a mouthful, but the concept is straightforward: Instead of predicting just one label, the model must choose many, sometimes dozens, from a set of thousands. Each dataset may need to be tagged with multiple, nuanced descriptors pulled from a controlled vocabulary.
Think of it like trying to identify all the animals in a photograph. If there’s just a dog, it’s easy. But if there’s a dog, a bird, a raccoon hiding behind a bush, and a unicorn that only shows up in 0.1% of your training photos, the task becomes far more difficult. That’s what GKR is up against: tagging complex datasets with precision, even when examples of some keywords are scarce.
And the problem is only growing. The new version of GKR now considers more than 3,200 keywords, up from about 430 in its earlier iteration. That’s a sevenfold increase in vocabulary complexity, and a major leap in what the model needs to learn and predict.
To handle this scale, the GKR team didn’t just add more data; they built a more capable model from the ground up. At the heart of the upgrade is INDUS, an advanced language model trained on a staggering 66 billion words drawn from scientific literature across disciplines—Earth science, biological sciences, astronomy, and more.
NASA ODSI’s GCMD Keyword Recommender AI model automatically tags scientific datasets with the help of INDUS, a large language model trained on NASA scientific publications across the disciplines of astrophysics, biological and physical sciences, Earth science, heliophysics, and planetary science. NASA “We’re at the frontier of cutting-edge artificial intelligence and machine learning for science,” said Sajil Awale, a member of the NASA ODSI AI team at MSFC. “This problem domain is interesting, and challenging, because it’s an extreme classification problem where the model needs to differentiate even very similar keywords/tags based on small variations of context. It’s exciting to see how we have leveraged INDUS to build this GKR model because it is designed and trained for scientific domains. There are opportunities to improve INDUS for future uses.”
This means that the new GKR isn’t just guessing based on word similarities; it understands the context in which keywords appear. It’s the difference between a model knowing that “precipitation” might relate to weather versus recognizing when it means a climate variable in satellite data.
And while the older model was trained on only 2,000 metadata records, the new version had access to a much richer dataset of more than 43,000 records from NASA’s Common Metadata Repository. That increased exposure helps the model make more accurate predictions.
The Common Metadata Repository is the backend behind the following data search and discovery services:
Earthdata Search International Data Network Learning to Love Rare Words
One of the biggest hurdles in a task like this is class imbalance. Some keywords appear frequently; others might show up just a handful of times. Traditional machine learning approaches, like cross-entropy loss, which was used initially to train the model, tend to favor the easy, common labels, and neglect the rare ones.
To solve this, NASA’s team turned to focal loss, a strategy that reduces the model’s attention to obvious examples and shifts focus toward the harder, underrepresented cases.
The result? A model that performs better across the board, especially on the keywords that matter most to specialists searching for niche datasets.
From Metadata to Mission
Ultimately, science depends not only on collecting data, but on making that data usable and discoverable. The updated GKR tool is a quiet but critical part of that mission. By bringing powerful AI to the task of metadata tagging, it helps ensure that the flood of Earth observation data pouring in from satellites and instruments around the globe doesn’t get lost in translation.
In a world awash with data, tools like GKR help researchers find the signal in the noise and turn information into insight.
Beyond powering GKR, the INDUS large language model is also enabling innovation across other NASA SMD projects. For example, INDUS supports the Science Discovery Engine by helping automate metadata curation and improving the relevancy ranking of search results.The diverse applications reflect INDUS’s growing role as a foundational AI capability for SMD.
The INDUS large language model is funded by the Office of the Chief Science Data Officer within NASA’s Science Mission Directorate at NASA Headquarters in Washington. The Office of the Chief Science Data Officer advances scientific discovery through innovative applications and partnerships in data science, advanced analytics, and artificial intelligence.
Share
Details
Last Updated Jul 09, 2025 Related Terms
Science & Research Artificial Intelligence (AI) Explore More
2 min read Polar Tourists Give Positive Reviews to NASA Citizen Science in Antarctica
Article
6 hours ago
2 min read Hubble Observations Give “Missing” Globular Cluster Time to Shine
Article
6 days ago
5 min read How NASA’s SPHEREx Mission Will Share Its All-Sky Map With the World
Article
7 days ago
Keep Exploring Discover Related Topics
Missions
Humans in Space
Climate Change
Solar System
View the full article
-
By USH
The photograph was captured by the Mast Camera (Mastcam) aboard NASA’s Curiosity rover on Sol 3551 (August 2, 2022, at 20:43:28 UTC).
What stands out in the image are two objects, that appear strikingly out of place amid the natural Martian landscape of rocks and boulders. Their sharp edges, right angles, flat surfaces, and geometric symmetry suggest they may have been shaped by advanced cutting tools rather than natural erosion.
Could these ancient remnants be part of a destroyed structure or sculpture? If so, they may serve as yet another piece of evidence pointing to the possibility that Mars was once home to an intelligent civilization, perhaps even the advanced humanoid beings who, according to some theories, fled the catastrophic destruction of planet Maldek and sought refuge on the Red Planet.
Objects discovered by Jean Ward Watch Jean Ward's YouTube video on this topic: HereSee original NASA source: Here
View the full article
-
By NASA
Curiosity Navigation Curiosity Home Mission Overview Where is Curiosity? Mission Updates Science Overview Instruments Highlights Exploration Goals News and Features Multimedia Curiosity Raw Images Images Videos Audio Mosaics More Resources Mars Missions Mars Sample Return Mars Perseverance Rover Mars Curiosity Rover MAVEN Mars Reconnaissance Orbiter Mars Odyssey More Mars Missions Mars Home 3 min read
Sols 4559-4560: Drill Campaign — Searching for a Boxwork Bedrock Drill Site
NASA’s Mars rover Curiosity acquired this image of a portion of its workspace, full of interesting but not drillable bedrock, using its Left Navigation Camera on June 2, 2025 — Sol 4558, or Martian day 4,558 of the Mars Science Laboratory mission — at 12:23:24 UTC. NASA/JPL-Caltech Written by Lucy Lim, Planetary Scientist at NASA’s Goddard Space Flight Center
Earth planning date: Monday, June 2, 2025
Now that Curiosity has spent a few sols collecting close-up measurements of the rocks in the outer edge of the boxwork-forming geologic unit, the team has decided that it’s time to collect a drill sample. The geochemical measurements by APXS and ChemCam have shown changes since we crossed over from the previous layered sulfate unit, but we can’t figure out the mineralogy from those data alone. As we’ve often seen before on Mars, the same chemical elements can crystallize into a number of different mineral assemblages. That’s even more the case in sedimentary rocks such as we are driving through, in which different grains in our rocks may have formed in different times and places. This also means that when we do get our mineral data, those minerals will tell us a lot about the history of these new-to-us rocks.
On board Curiosity, that mineral analysis is the job of the CheMin instrument, which uses X-ray diffraction to identify minerals. CheMin shines a narrow X-ray beam through a powdered sample in order to generate the diffraction pattern, which means that it needs a drilled sample. So the team today was busy looking for a drillable spot. Unfortunately the rover’s drill reach from today’s parking spot included only rocks that were too fractured or had too much debris sitting on them to be considered likely to produce a good drilled sample, so we will have to move, or “bump,” at least one more time before progressing to the drill preload test, which is the next step in drilling.
In the meantime, we are taking more measurements to understand the range of compositions that can be found in this rock layer. Dust removal (DRT) + APXS + LIBS + MAHLI were all planned for target “Holcomb Valley,” while a short distance away a second DRT/APXS/MAHLI measurement was planned for “Santa Ysabel Valley” and in another direction, a second LIBS for “Stough Saddle.” One long-distance ChemCam remote imaging mosaic was planned to cover a boxwork structure off in the distance. Mastcam had a relatively light day of imaging, with just a couple of small mosaics covering a nearby trough feature, and providing context for the RMI of the boxwork structure, in addition to documenting the two LIBS targets. The modern Mars environment was also recorded with a couple of movies to look for dust-devil activity, a measurement of atmospheric opacity, and a pair of suprahorizon observations to look for clouds, plus the usual passive observations by DAN and REMS to monitor the neutron environment, temperature, and humidity.
I’ll be on rover planning Wednesday as Geology and Mineralogy Science Theme Lead and looking forward to what we find — hopefully some drillable boxwork-unit bedrock!
Share
Details
Last Updated Jun 04, 2025 Related Terms
Blogs Explore More
2 min read Sols 4556-4558: It’s All in a Day’s (box)Work
Article
1 day ago
2 min read Sols 4554–4555: Let’s Try That One Again…
Article
6 days ago
2 min read Sol 4553: Back to the Boxwork!
Article
6 days ago
Keep Exploring Discover More Topics From NASA
Mars
Mars is the fourth planet from the Sun, and the seventh largest. It’s the only planet we know of inhabited…
All Mars Resources
Explore this collection of Mars images, videos, resources, PDFs, and toolkits. Discover valuable content designed to inform, educate, and inspire,…
Rover Basics
Each robotic explorer sent to the Red Planet has its own unique capabilities driven by science. Many attributes of a…
Mars Exploration: Science Goals
The key to understanding the past, present or future potential for life on Mars can be found in NASA’s four…
View the full article
-
By USH
What would you do if you suddenly felt an unseen presence, turned around—and found yourself face to face with a seven-foot-tall, insect-like entity? Since 2006, anglers along New Jersey’s Musconetcong River have reported startling encounters with just such a being: a towering, humanoid creature that closely resembles a praying mantis.
But these aren’t just fleeting sightings. Witnesses frequently describe deeply unsettling experiences: telepathic communication, a sense of their thoughts or memories being accessed, and profound physiological effects. Consistent patterns emerge—electronic devices glitch, the surrounding forest falls unnaturally silent, and a strange, low-frequency hum seems to vibrate through the air.
More intriguingly, these mantis-like figures aren’t limited to modern encounters. Strikingly similar forms appear in ancient art across the globe, from 8,000-year-old cave paintings to references in Egyptian iconography. Could these entities have been with us since the dawn of civilization?
Theories vary widely. Some suggest these beings are an advanced species of insectoid extraterrestrials, possibly master geneticists overseeing hybridization programs involving humanity. Others propose a more Earth-bound origin, perhaps they’re a secret lineage of evolved terrestrial insects, hiding in the shadows of time.
And then there’s the interdimensional hypothesis: that these creatures aren’t physical in the way we understand, but exist in a parallel state of reality, occasionally phasing into ours.
Some researchers have even speculated that geological fault lines, like those beneath the Musconetcong River, could serve as energetic gateways, allowing these entities to cross between dimensions.
One thing is clear: the Mantis beings are watching and they may have been here far longer than we’ve dared to imagine.
View the full article
-
By NASA
Explore This Section Perseverance Home Mission Overview Rover Components Mars Rock Samples Where is Perseverance? Ingenuity Mars Helicopter Mission Updates Science Overview Objectives Instruments Highlights Exploration Goals News and Features Multimedia Perseverance Raw Images Images Videos Audio More Resources Mars Missions Mars Sample Return Mars Perseverance Rover Mars Curiosity Rover MAVEN Mars Reconnaissance Orbiter Mars Odyssey More Mars Missions Mars Home 2 min read
Searching for Spherules to Sample
Subsurface spherules: This image of the Hare Bay abrasion patch was acquired by the WATSON camera on Sol 1480 (April 19, 2025), showing dark-colored spherules set in a fine-grained light-toned matrix. These spherules appear to be smaller versions of similar structures that have been found in numerous rocks in the vicinity. Perseverance is currently working to collect a sample of these spherules to return to Earth. WATSON (Wide Angle Topographic Sensor for Operations and eNgineering) is a close-range color camera that works with the rover’s SHERLOC instrument (Scanning Habitable Environments with Raman and Luminescence for Organics and Chemicals); both are located on the turret at the end of the rover’s robotic arm. NASA/JPL-Caltech Written by Denise Buckner, Postdoctoral Fellow at NASA’s Goddard Space Flight Center
Over the past few weeks, Perseverance has been investigating some curious spherules peppered across the “Witch Hazel Hill” region along the rim of Jezero crater. A striking cluster of the small bubble-shaped stones were first spotted by the Mastcam-Z instrument on Sol 1442 (March 11, 2025) at “Broom Point,” in a rock named “St. Pauls Bay.” A few sols later, a similar assemblage was discovered by the SuperCam instrument at the “Mattie Mitchell” outcrop near “Puncheon Rock.” As the rover continued along its traverse, spherules continued to appear. At the targets St. Pauls Bay and Mattie Mitchell, the spherules are densely packed and almost look like bunches of grapes. Elsewhere, similar smaller spherules were found intermixed with other grains within the rock. At a target called “Wreck Apple” at the “Sally’s Cove” outcrop, individual spherules were set in a matrix of coarse, dark grains. Even more of these circular features are embedded in finer-grained, layered bedrock at a nearby area called “Dennis Pond.”
Spherules at St. Pauls Bay: NASA’s Mars Perseverance rover acquired this image, a striking cluster of spherules, on March 11, 2025 – Sol 1442, or Martian day 1,442 of the Mars 2020 mission – at the local mean solar time of 11:12:40. Perseverance used its Left Mastcam-Z camera; Mastcam-Z is a pair of cameras located high on the rover’s mast. NASA/JPL-Caltech/ASU Spherules at Wreck Apple: NASA’s Mars Perseverance rover found smaller spherules in a coarse-grained matrix. The rover captured this image using the WATSON camera on March 27, 2025 – Sol 1458, or Martian day 1,458 of the Mars 2020 mission – at the local mean solar time of 15:36:04. WATSON (Wide Angle Topographic Sensor for Operations and eNgineering) is a close-range color camera located on the turret at the end of the rover’s robotic arm. NASA/JPL-Caltech Although the team was intrigued by the spherule-rich layers at Sally’s Cove and Dennis Pond, these outcrops were challenging for the rover arm to access. After some searching to find an accessible target, the team decided to perform an abrasion at a neighboring outcrop, called “Pine Pond,” which contained an extension of the Dennis Pond layers. The team picked the target “Hare Bay” in hopes of finding spherules within a rock interior, and conducting proximity science observations with PIXL and SHERLOC to investigate their composition and internal structure. Images of the abrasion patch taken by WATSON show that Hare Bay contains light-toned medium-sized grains, with millimeter-sized spherules dotted throughout the rock! Leading hypotheses for the origin of these spherules include formation by volcanic activity or impact-related processes.
Having found an accessible spherule-bearing rock, the team is currently hard at work collecting a spherule-filled sample! Combined with the information already gathered by Mastcam-Z, SuperCam, PIXL, SHERLOC, and WATSON, future laboratory analyses could help solve the mystery of when, where, and how these spherules formed, which can in turn detangle the geological events that formed and transformed the surface of Mars over billions of years!
Share
Details
Last Updated May 05, 2025 Related Terms
Blogs Explore More
2 min read Sols 4527-4528: ‘Boxwork Ahoy!’
Article
22 hours ago
3 min read Sols 4525-4526: The Day After Groundhog Day (Between Ghost Mountain and Texoli, Headed South)
Article
5 days ago
4 min read Sols 4522-4524: Up on the Roof
Article
6 days ago
Keep Exploring Discover More Topics From NASA
Mars
Mars is the fourth planet from the Sun, and the seventh largest. It’s the only planet we know of inhabited…
All Mars Resources
Explore this collection of Mars images, videos, resources, PDFs, and toolkits. Discover valuable content designed to inform, educate, and inspire,…
Rover Basics
Each robotic explorer sent to the Red Planet has its own unique capabilities driven by science. Many attributes of a…
Mars Exploration: Science Goals
The key to understanding the past, present or future potential for life on Mars can be found in NASA’s four…
View the full article
-
-
Check out these Videos
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.