Google has added some new security features to Gmail, and one of them is reportedly improving the detection of phishing or Phishing χρησιμοποιώντας σε ένα εξειδικευμένο μοντέλο engineeringof learning.
According to the company 50 up to 70 percent of incoming messages in gmail is spam, but learning engineering techniques help Google beat spam and phishing messages with accuracy that reaches 99,9%.
The latest engineering learning model improves the process by delaying selected messages (less than the average of 0,05% of messages) for further fishing analysis.
Google reports that new crawling patterns will also help create new warnings delaying time accesss in the URL, showing users a warning question when they have clicked on a suspicious link. As new standards are identified, new protection models will adapt and improve over time.
In addition, Google added external company alert warnings to prevent users from sending accidentally protected data to someone outside of their company.
Gmail will not stop sending, but with warnings it will be trying to delay it a bit, according to Google, as mechanical intelligence will alert Gmail if the recipient is an existing contact or someone with whom the sender interacts regularly.
Google also added new built-in defenses against ransomware and polymorphic malware software, blocking millions of add-ons that circulate as email attachments.
Threats are identified by combining flags of spam, malware, and ransomware with euristics attachments (emails that may be flagship threats), and digital signatures of senders (already identified as malicious programs).
