Algorithmic Selection on the Internet

The Internet is increasingly permeating daily life with an essential and structuring bundle of innovations: Internet-based applications that operate on algorithmic selection. Automated algorithmic selection is embedded in a variety of services and applied for numerous purposes. Algorithms in search engines or news aggregators select information in general and news in particular; in recommender systems, they suggest music and video entertainment or influence one’s choice of products. They may also suggest friends, partners, and travel routes. Moreover, they are used to automatically produce news articles and messages in social media, to calculate scoring of content and people, and to observe behavior and interests as well as to predict and shape future needs and actions. Although their modes of operation differ in detail, all these applications share a common defining functionality: They automatically select information elements and assign relevance to them.

In information societies subject to growing datafication, Internet-based algorithmic selection is indispensable for extracting social and economic values from big and small data. Accordingly, these automatic assignments of relevance to selected pieces of information are already deeply embedded in many domains of everyday life. They shape how the world is perceived, what realities are constructed, and how people behave. Their comprehensive adoption has consequently raised discussions about opportunities and social risks, such as surveillance, echo chambers, filter bubbles, and social discrimination, as well as questions regarding appropriate public policy and governance measures.

This entry gives a brief overview of research on the topic of algorithmic selection on the Internet and offers a functional typology of applications operating on it. It then addresses their economic significance, social risks, and governance options.