Two clicks to cluster and block thousands of spammers
summary
LOVOO uses unsupervised models to let moderators two clicks cluster and act on user batches; the platform stitches an ML model service into an Antispam system with Kafka and gRPC, databases and feature pipelines. It identifies thousands of likely bad accounts, for example a cluster of 340 with ~90% spammers, and supports bulk actions and audit trails. Deployment reduces manual review by surfacing high-confidence clusters and enabling fast, reproducible blocks.
AI-written summary · about
chapters (5)
- 0:00Introduction to spammer blocking
- 5:00Clustering spammers in practice
- 15:00Real-world results and metrics
- 25:00Optimizing the system further
- 30:00Q&A and closing remarks
related talks
16mProduction MLOps for Spam Detection: Real-Time Multi-Modal AI
30mML Automoderation System for LLM-Augmented Online Platforms
33mMachine Learning to moderate ads
24mHighly Scalable Discriminative Spam Filtering
42mOnline Safety Bill: How Global Platforms Use MLOps to Keep People Safe
29mSocial Media Botnet Detection with Linkage Analysis and Machine Learning
19mMLOps End-to-End Speedrun: Build an APIs around a Spam Classifier Model
59mCalculating by IP: How to Fight Spam on a Social Network