The burst of the bubble? Effects of automated personalization on news diversity
- insert_drive_file Peer-Reviewed Presentations
- event 2016
- translate English
Modern search engines and news websites increasingly use algorithms to personalize contents to individual recipients. Personalized content might confront people with limited news offers and result in so-called "filter bubbles". Filter bubbles are said to threaten news diversity on two dimensions – topic and source diversity. Two studies look into the degree of "bubble formation" of Google search results. In the first study, we created four different Google accounts. Each account was "trained" over a one-week period to simulate one of four media usage types. In a subsequent test-period we found no difference in Google search results across the four accounts. In the second study, we looked at Google News entries to analyze source diversity. Results indicate that there is no filter bubble in terms of online news topics. Yet, looking at source diversity, Google news over-represents some news providers, leading to a biased spectrum of news sources. We conclude that current concerns about "filter bubbles" are overstated.
Haim, M., Graefe, A., & Brosius, H.-B. (6/2016). The burst of the bubble? Effects of automated personalization on news diversity. Presented at the 66th Annual Conference of the International Communication Association, Fukuoka. (content_copy)