Skip to main content
  1. Home
  2. Phones
  3. Android
  4. Mobile
  5. News

Here’s how the Google Play Store detects fake ratings and reviews

Add as a preferred source on Google
Google Play Store on Huawei P20 Pro
Simon Hill/Digital Trends

There are tons of options on the Google Play Store, whether you’re looking for a sleep tracker, yoga guide, or a new game. If you’re anything like us, each app’s ratings and reviews are crucial to deciding which app to download — and so it’s especially bitter to learn that many of those ratings or reviews are faked. That’s why Google is working hard to uncover as many fake and unhelpful reviews and malevolent apps as possible.

What comprises the material that Google is so keen to get rid of? Simply put, it’s content that’s either not helpful, profane, or simply not truthful. Google classes bad reviews into three separate categories: “Bad content” — reviews which are profane, off-topic, or hateful. “Fake reviews” are any review which seeks to unduly influence an app’s position in ratings, either by inflating the number of positive reviews, or tanking another app’s reviews with false low ratings. And “incentivized reviews” come from genuine users, but those who have been offered money or valuable items in-app in exchange for a glowing review.

An example of an app lending favors for a positive review Image used with permission by copyright holder

To help fight this tide of fake news reviews, 2018 saw Google deploy a new system that combines human intelligence with machine learning to detect and enforce policy violations. Engineers and analysts keep an eye on the Play Store’s review trends to make sure the machine learning model is kept up to date, and skilled reviewers are often brought in to check the decisions made by the review-checking models.

Recommended Videos

It sounds complicated, but it clearly works. According to Google’s post, a recent week saw millions of bad reviews and rating removed from the Play Store, and thousands of bad apps were also given the boot, detected in large part by those same false reviews. Despite a security crackdown in the last few years, malicious apps are still finding their way onto the Play Store — as the recent fake cryptocurrency apps can attest.

So Google seems to be going gangbusters when it comes to cracking down on false reviews — but it has also admitted that it still needs help from the community. If you spot a review that breaks Google’s guidelines on profanity or hate, or is simply off-topic, you can flag it as “Spam,” or mark it as “Inappropriate” from the review’s options. Developers can instead submit feedback on specific comments through the Play Console.

Mark Jansen
Former Mobile Evergreen Editor
Mark Jansen is an avid follower of everything that beeps, bloops, or makes pretty lights. He has a degree in Ancient &…
Pixi wants to replace your boring text messages with AR characters that react to you
iMessage users can now send fun AI characters like a cat or robot to their friends.
pixi-ar-app-imessage

Forget stickers and GIFs, a new app called Pixi Garden wants you to send interactive augmented reality characters through iMessage instead.

Pixi Platforms launched the messaging native app today, letting you create and send a "pixi" — an intelligent AR character that comes alive through your friend's phone camera and reacts to whatever is actually happening around them.

Read more
AI vision is getting too hungry, and this method puts it on a diet
KAIST researchers say Upsample Anything sharpens compressed visual data while cutting GPU memory demands by up to 16 times.
Car, Transportation, Vehicle

KAIST researchers have developed an AI vision method built for a problem phone makers can’t ignore forever. Upsample Anything rebuilds high-resolution visual features from compressed image data, aiming to make on-device AI sharper without demanding a much bigger memory budget.

Phones already lean on compression to keep camera-based intelligence moving quickly. The tradeoff is that small objects, thin edges, and subtle defects can get stripped away before a vision system has enough detail to work with.

Read more
Google Photos’ AI image editor expands to more regions, but only for Android users
Edit with Ask Photos, which lets you make edits by describing what you want, is now available for Android users in Germany, the UK, France, Spain, and Italy.
Featured image for Google Photos conversational edit ability.

Google introduced an AI-powered editing feature in Google Photos called "Edit with Ask Photos" last year, allowing users to make photo adjustments using natural language prompts. It initially debuted in a handful of countries, but Google is now expanding support to five new markets.

From four countries to nine

Read more