Programs corresponding to Gmail, YouTube and Google Play depend on textual content classification fashions to determine dangerous content material together with phishing assaults, inappropriate feedback, and scams. All these texts are more durable for machine studying fashions to categorise as a result of dangerous actors depend on adversarial textual content manipulations to actively try to evade the classifiers. For instance, they are going to use homoglyphs, invisible characters, and key phrase stuffing to bypass defenses.
To assist make textual content classifiers extra strong and environment friendly, we’ve developed a novel, multilingual textual content vectorizer referred to as RETVec (Resilient & Environment friendly Textual content Vectorizer) that helps fashions obtain state-of-the-art classification efficiency and drastically reduces computational value. At the moment, we’re sharing how RETVec has been used to assist defend Gmail inboxes.
Strengthening the Gmail Spam Classifier with RETVec
Determine 1. RETVec-based Gmail Spam filter enhancements.
Over the previous 12 months, we battle-tested RETVec extensively inside Google to judge its usefulness and located it to be extremely efficient for safety and anti-abuse purposes. Particularly, changing the Gmail spam classifier’s earlier textual content vectorizer with RETVec allowed us to enhance the spam detection charge over the baseline by 38% and scale back the false constructive charge by 19.4%. Moreover, utilizing RETVec lowered the TPU utilization of the mannequin by 83%, making the RETVec deployment one of many largest protection upgrades lately. RETVec achieves these enhancements by sporting a really light-weight phrase embedding mannequin (~200k parameters), permitting us to scale back the Transformer mannequin’s measurement at equal or higher efficiency, and being able to separate the computation between the host and TPU in a community and reminiscence environment friendly method.
RETVec Advantages
RETVec achieves these enhancements by combining a novel, highly-compact character encoder, an augmentation-driven coaching regime, and using metric studying. The structure particulars and benchmark evaluations can be found in our NeurIPS 2023 paper and we open-source RETVec on Github.
As a consequence of its novel structure, RETVec works out-of-the-box on each language and all UTF-8 characters with out the necessity for textual content preprocessing, making it the perfect candidate for on-device, internet, and large-scale textual content classification deployments. Fashions educated with RETVec exhibit sooner inference pace on account of its compact illustration. Having smaller fashions reduces computational prices and reduces latency, which is crucial for large-scale purposes and on-device fashions.
Determine 1. RETVec structure diagram.
Fashions educated with RETVec could be seamlessly transformed to TFLite for cellular and edge gadgets, because of a local implementation in TensorFlow Textual content. For internet utility mannequin deployment, we offer a TensorflowJS layer implementation that’s obtainable on Github and you’ll take a look at a demo internet web page working a RETVec-based mannequin.
Determine 2. Typo resilience of textual content classification fashions educated from scratch utilizing totally different vectorizers.
RETVec is a novel open-source textual content vectorizer that means that you can construct extra resilient and environment friendly server-side and on-device textual content classifiers. The Gmail spam filter makes use of it to assist defend Gmail inboxes in opposition to malicious emails.
If you need to make use of RETVec in your personal use instances or analysis, we created a tutorial that can assist you get began.
This analysis was carried out by Elie Bursztein, Marina Zhang, Owen Vallis, Xinyu Jia, and Alexey Kurakin. We want to thank Gengxin Miao, Brunno Attorre, Venkat Sreepati, Lidor Avigad, Dan Givol, Rishabh Seth and Melvin Montenegro and all of the Googlers who contributed to the venture.
