Researchers have developed a search engine tool that shows what Google search results appear in different countries or languages, highlighting key differences in the algorithm between regions.
Search Atlas allows users to cross the borders of the "filter bubble" they live in, which Google creates using data on a person's location, language, search history. The tool was created as part of a study conducted by two doctoral students at Carnegie Mellon University and MIT.
The Google search algorithm already ranks and prioritizes certain websites in search results, whether motivated by the designers’ academic, political or financial interests. The creators of Search Atlas invite users to "reflect on how their online lives are conditioned by technological infrastructures and geopolitical regimes.”
“Search engines both reflect the world and remap it, determining the information that users see. We built Search Atlas to let you search beyond borders,” Katherine Ye, a doctoral student at Carnegie Mellon University who co-created the tool, recently tweeted.
Ye and Rodrigo Ochigame, a doctoral student from MIT, created Search Atlas using a third party data scraper to collect search data from Google. Viewing the side-by-side differences in search results shows how Google controls what kind of information users can see.
Google’s massive data index makes it the first tool people turn to when trying to learn something new. “Its dominance is so unquestioned and even people who are thinking critically about technologies are ignoring this big elephant in the room that is search,” Ye told Motherboard.
In one instance, the study references how the region of Kashmir is depicted with a solid border on Google Maps for users in India. But it appears with a dotted border near Pakistan when it’s searched in other countries, showing that Kashmir’s ownership is disputed.
Ochigame and Ye used Search Atlas to search for specific topics that could show the greatest geopolitical differences, such as God or how to combat climate change. A search for Tiananmen Square in the United Kingdom and Singapore pulled up images of soldiers and the famous Tank Man image, but a search in China only shared promotional, tourist images of Tiananmen Square.
Ye described Google as a giant scraper, meaning it generates search results by scraping data from every single webpage to create its massive information index. But Google makes it extremely hard to research or audit its own data.
The company's scale and monopoly on data also makes it so that there are few other options available for search engines outside of Google. As a possible alternative, Ye referenced DuckDuckGo, which does not collect users’ private data, but with a scale and level of efficiency that is considerably small compared to Google.
“This is really helping people broaden their field of view,” said Ye. “It’s just a very small first step or one of several first steps, including building on work that other people have done, toward exposing this power, towards challenging it, building public knowledge and perception of how powerful these technological infrastructures are.”