The researchers developed a system that can identify when people's online behavior suddenly changes in ways that might be a threat to their private information. This new approach could potentially help protect people's personal data by flagging suspicious activity, and it has the potential to become an important tool for safeguarding online privacy.
Read Full Paper →The team created a new system using machine learning that can quickly spot when devices connected to the internet are under attack. This could help prevent many types of cyberattacks, which have been on the rise as more "smart" devices like home appliances and security cameras connect to the web.
Read Paper →Scientists analyzed how attacks on important infrastructure like power plants and water treatment facilities happen, using data from past cyberattacks in Iran and Ukraine. Their findings show that even when these systems are not connected to other networks, they can still be vulnerable to attack due to weaknesses within their own control systems.
Read Paper →Scientists created a new system that helps protect private documents stored online in India's digital storage service, called DigiLocker. This improved protection could help people feel more confident storing sensitive information online, and it could also give the Indian government more ways to provide secure services to its citizens.
Read Paper →Scientists created a new technique called Proportional Progressive Pseudo Labeling (PPPL) to help computers learn from one type of data and apply it to another, even if they look or behave very differently. This technique has the potential to improve how well computers can perform tasks such as recognizing images, analyzing text, and detecting unusual patterns in data.
Read Paper →
