Knowledge sharing and Yahoo Answers: Everyone knows something

Abstract

Yahoo Answers (YA) is a large and diverse question-answer
\nforum, acting not only as a medium for sharing technical
\nknowledge, but as a place where one can seek advice, gather
\nopinions, and satisfy one's curiosity about a countless number
\nof things. In this paper, we seek to understand YA's
\nknowledge sharing activity. We analyze the forum categories
\nand cluster them according to content characteristics and
\npatterns of interaction among the users. While interactions
\nin some categories resemble expertise sharing forums, others
\nincorporate discussion, everyday advice, and support. With
\nsuch a diversity of categories in which one can participate,
\nwe nd that some users focus narrowly on speci c topics,
\nwhile others participate across categories. This not only allows
\nus to map related categories, but to characterize the
\nentropy of the users' interests. We nd that lower entropy
\ncorrelates with receiving higher answer ratings, but only for
\ncategories where factual expertise is primarily sought after.
\nWe combine both user attributes and answer characteristics
\nto predict, within a given category, whether a particular answer
\nwill be chosen as the best answer by the asker.