This week’s list of top data news highlights covers July 9, 2022 to July 15, 2022 and includes articles on creating smart jean pockets and verifying information found online.
1. Sorting Through Recorded Evidence
The Virginia Indigent Defense Commission, a state agency that provides criminal defense representation to indigent individuals, has announced plans to use a machine learning model to sort through recordings from police body cameras, interrogations, and 911 calls. The model creates transcripts from the footage and can flag relevant parts such as mentions of drug use or weapons for public defenders.
An international team of researchers has identified the original location of a 4.48 billion-year-old Martian meteorite. The team used a supercomputer to match the meteorite’s compositional characteristics with known Martian craters and found that the meteorite was likely ejected during the formation of a crater in the Terra Cimmeria-Sirenum region of Mars.
3. Verifying Wikipedia Entries
Meta has created an AI system that can verify citations in Wikipedia articles. Researchers trained the system with 4 million Wikipedia citations. The system uses data from 134 million public webpages to ensure citations support the information in entries.
Microsoft has launched the Digital Equity Data Dashboard to show equity gaps in access to digital resources across the United States. The dashboard uses data such as broadband access, education, and poverty rates from the Census Bureau, Federal Communications Commission, BroadbandNow, and Microsoft’s Broadband Usage Data to show digital equity by census tract.
5. Predicting Pain Management Issues
Researchers at Cedars-Sinai Medical Center, a U.S.-based healthcare organization, have created an AI system that can predict which patients can manage their pain after surgeries and which patients will require assistance with pain management. Doctors can use the system to determine patients’ risk of medication misuse.
Amazon is deploying smart shopping carts at a Whole Foods store in Massachusetts. Each cart contains sensors and an AI system that can identify which groceries shoppers place in or remove from the cart as well as an interactive touch screen showing the location of nearby groceries.
Meta has created an AI system that can transform user-submitted text and sketches into images. Their system enables more specificity than previous systems by allowing users to submit simple sketches as well as text.
Microsoft has created a smart fabric, designed for jeans’ pockets, that can recognize objects and gestures. The fabric uses sensors and a machine learning model to recognize the presence or absence of common objects.
9. Identifying Drug Candidates
Researchers at the Massachusetts Institute of Technology have created an AI system that can identify drug molecule candidates by predicting how proteins will bind to the molecule. In tests, the system identified promising drug molecules 1,200 times faster than current methods.
Officials in Edinburgh, United Kingdom have started installing 11,000 smart waste bins around the city. Each bin contains a sensor that can alert officials when it is full and collect data on usage. Officials can use this data to create more efficient waste collection routes and management plans.
Image credit: Flickr user pepperberry farm