Abstract
Web-scale discovery has arrived. With products like Summon and WorldCat Local, hundreds of millions of articles and books are accessible at lightning speed from a single search box via the library. But there's a catch. As the size of the index grows, so too does the challenge of relevancy. When Google launched in 1998 with an index of only 25 million pages, its patented PageRank algorithm was powerful enough to provide outstanding results. But the web has grown to well over a trillion pages, and Google now employs over 200 different signals to determine what search results you see. According to Eli Pariser, author of "The filter bubble: what the internet is hiding from you" (Penguin, 2011), a growing number of these signals are based on what Google knows about you, especially your web history; and, according to Pariser, serving up information that's "pleasant and familiar and confirms your beliefs" is becoming increasingly synonymous with relevancy. This session will critique Pariser's concept of the 'filter bubble' in terms of collection development and the possible evolutions of discovery layers like Summon and WorldCat Local, and the challenge of providing relevant academic research results in a web-scale world where students increasingly expect the kind of personalization sometimes at odds with academia's adherence to privacy and intellectual freedom.
Relevancy Redacted: Web-Scale Discovery and the “Filter Bubble”
Web-scale discovery has arrived. With products like Summon and WorldCat Local, hundreds of millions of articles and books are accessible at lightning speed from a single search box via the library. But there's a catch. As the size of the index grows, so too does the challenge of relevancy. When Google launched in 1998 with an index of only 25 million pages, its patented PageRank algorithm was powerful enough to provide outstanding results. But the web has grown to well over a trillion pages, and Google now employs over 200 different signals to determine what search results you see. According to Eli Pariser, author of "The filter bubble: what the internet is hiding from you" (Penguin, 2011), a growing number of these signals are based on what Google knows about you, especially your web history; and, according to Pariser, serving up information that's "pleasant and familiar and confirms your beliefs" is becoming increasingly synonymous with relevancy. This session will critique Pariser's concept of the 'filter bubble' in terms of collection development and the possible evolutions of discovery layers like Summon and WorldCat Local, and the challenge of providing relevant academic research results in a web-scale world where students increasingly expect the kind of personalization sometimes at odds with academia's adherence to privacy and intellectual freedom.