Search & Filter Bubble

Overview
A Filter or Search Bubble is a phenomenon that occurs when personalized search results cause intellectual isolation and create a filtered mental bubble. Search engine ranking algorithms display results on the basis of not only individual user queries, a data collection of user queries, and page structure and elements, but also on the basis of personal information obtained from users. Past queries attained from users’ browsing or search history and user location are types of information perhaps unknowingly released by users. Data mining techniques are utilized to selectively filter out results based on this personal information, only displaying results that are highly relevant to an individual, effectively filtering out the influence of information outside of users’ comfort zone.

Personalized web results are altering the way internet users think on both personal and cultural levels. Search algorithms guess what each individual prefers to see displayed in their results, eliminating opposing views from reaching their eyes.

Intellectual Isolation from the Search Bubble
As a result of selectively displayed information from search engines, people are not presented with information that disagrees with their ways of thinking. Users become accustomed to not having to engage with ideas that conflict with what they understand. Ultimately, this isolates the user, making it more challenging to break through their own cultural barriers and understanding. There is debate on whether this type of over personalization is ethical. Luckily, there are ways for users to counteract the filter bubble phenomenon by relying on themselves and features within these technologies.

Countermeasures on the Individual Level
Individuals can counteract this phenomenon, for instance, by clearing their browsing history, utilizing the anonymous web browsing feature offered by their browsers. User search history offers a lot of data for search engines to use to filter out results to fit users’ current perspectives. Using anonymous browsers prevents search engines from utilizing this information when it comes to user search results. However, data from private browsing can be pretty easily recovered, but this would not play a role in individualized search results. Users can also disable and delete browser or HTTP cookies, which is data collected by web browsers when users visit a site. When permission for cookies is disabled, this data cannot be communicated across different sites.

To reduce location-based tracking, users can choose to utilize a virtual private network or a VPN. VPNs conceal IP addresses and they use servers located around the world to conceal users’ locations.

Outside of controlling these technology-based aspects, users can take a personal level of control, where they must actively evaluate the information they encounter. Users can ensure that they are being exposed to a wide range of ideas and perspectives, to promote critical thinking.

Countermeasures on the Business Level
Companies, like Facebook and Google, can take steps towards removing filter bubbles from the internet and using other means to offer information to its users. Companies can do this by ensuring they are prioritizing content that is reputable and has varying perspectives on shared topics.

Maccatrozzo discusses the use of semantic web techniques to introduce, as a performance measure in algorithms, novel concepts that are relevant to user-search. Semantic-based topic recommendations can be made in conjunction with user profile modeling to promote the introduction of opposing views in search results.