Google Maps using ML to remove fake business and misleading images

Google Maps has over one billion active users every month. In a recent blog post, the navigation service explained how it is using machine learning to take down fake content on Google Maps even before it was posted. Google also shared an example where its machine learning algorithm detected a sudden spike in the number of websites ending with .design or .top.

Google says it is using a ‘ a new machine learning model that could recognize numbers overlaid on contributed images by analyzing specific visual details and the layouts of photos.’ For Google Maps, the new model seems to be working as it is taking down the majority of such content even before it is published.

The company also said that it has blocked over 115 million policy violating reviews in 2022 and 20 per cent more fake reviews compared to 2021. The new models also helped block or remove over 200 million photos and 7 million videos that were blurry or low quality or violated the policy. In 2022, Google also stopped more than 20 million attempts to create fake business profiles and ‘185,000 businesses after detecting suspicious activity and abuse attempts.’Last year, the tech giant made updates to its machine learning models to help identify novel abuse trends.