Google is working on AI-powered anti-spam tools to make Gmail safer
Described as a significant defense upgrade, the latest development from Google is a cutting-edge text classification system known as RETVec (Resilient and Efficient Text Vectorizer).
Google's company stated that it recently introduced a cutting-edge multilingual text vectorizer called RETVec. This innovative tool enhances the performance of text classifiers, enabling them to achieve top-notch classification results while significantly reducing computational requirements.
Systems like Gmail, YouTube, and Google Play heavily depend on text classification models to detect and combat harmful content, such as malware, inappropriate comments, and frauds.
As their tech expert said, various techniques are employed to evade defenses, such as utilizing homoglyphs, invisible characters, and keyword stuffing.
These kinds of texts pose a challenge for machine learning models to categorize due to the deliberate efforts of malicious individuals to manipulate the text and avoid detection by the classifiers.
Thanks to its unique design, RETVec is compatible with all languages and characters, eliminating the need for text preprocessing. This makes it a perfect choice for on-device, web, and large-scale text classification projects.
RETVec is an innovative open-source text vectorizer that empowers individuals to create robust and streamlined server-side and on-device text classifiers.
According to Google, the Gmail spam filter utilizes this technology to safeguard Gmail inboxes from harmful emails. https://economictimes.indiatimes.com/