Predictors of Answer Quality in Online strongQstrongA Sites.pdfVIP

Predictors of Answer Quality in Online QA Sites * ** ** * F. Maxwell Harper , Daphne Raban , Sheizaf Rafaeli , Joseph A. Konstan * ** GroupLens Research Center for the Study of the Information Society Department of Computer Science and Engineering Graduate School of Management University of Minnesota University of Haifa Minneapolis, Minnesota, USA Mount Carmel, Haifa, Israel {harper, konstan}@ draban@univ.haifa.ac.il; sheizaf@ ABSTRACT South Korea, dwarfing worldwide leaders Yahoo! (4.4%) Question and answer (QA) sites such as Yahoo! and Google (1.7%) [19]. One of the reasons for this Answers are places where users ask questions and others disparity is the relatively small Korean language corpus answer them. In this paper, we investigate predictors of available for crawling. To address this shortcoming, Naver answer quality through a comparative, controlled field built a QA site called Knowledge iN that encourages study of responses provided across several online QA users to type questions for others to answer, rather than sites. Along with several quantitative results concerning the