No abstract

No abstract

While Wikipedia is a subject of great interest in the computing literature, very little work has considered Wikipedia’s important relationships with other information technologies like search engines. In this paper, we report the results of two deception studies whose goal was to better understand the critical relationship between Wikipedia and Google. These studies silently removed Wikipedia content from Google search results and examined the effect of doing so on participants’ interactions with both websites. Our findings demonstrate and characterize an extensive interdependence between Wikipedia and Google. Google becomes a worse search engine for many queries when it cannot surface Wikipedia content (for example, click-through rates on results pages drop significantly) and the importance of Wikipedia content is likely greater than many improvements to search algorithms. Our results also highlight Google’s critical role in providing readership to Wikipedia. However, we also found evidence that this mutually beneficial relationship is in jeopardy: changes Google has made to its search results that involve directly surfacing Wikipedia content are significantly reducing traffic to Wikipedia. Overall, our findings argue that researchers and practitioners should give deeper consideration to the interdependence between peer production communities and the information technologies that use and surface their content.

No abstract

No abstract

No abstract

No abstract

Much research has shown that social media platforms have substantial population biases. However, very little is known about how these population biases affect the many algorithms that rely on social media data. Focusing on the case study of geolocation inference algorithms and their performance across the urban-rural spectrum, we establish that these algorithms exhibit significantly worse performance for underrepresented populations (i.e. rural users). We further establish that this finding is robust across both text-and network-based algorithm designs. However, we also show that some of this bias can be attributed to the design of algorithms themselves rather than population biases in the underlying data sources. For instance, in some cases, algorithms perform badly for rural users even when we substantially overcorrect for population biases by training exclusively on rural data. We discuss the implications of our findings for the design and study of social media-based algorithms.