Simple Tags, Profound Influences: Explicating Social Tagging as Multilevel Expertise in a Crowdsourcing Design Community

• In Event: HIGH-DENSITY: Exploring Information Systems: Models and Theories

Abstract

For designers in crowdsourcing business, tagging their artifacts is a deceptively simple task. What has been overlooked is how certain tags are “intuitively” selected, as well as the influences these words exert on the taggers’ performances. Using data from a major crowdsourcing design community over a seven-year period of time, the study showcases that sensing the “right” tags is a critical skill for a designer to succeed. By performing k-core algorithm on a bipartite network consisting of over 11,000 nodes, a multi-layered core-periphery structure was identified, showing that some tags gradually occupied more advantageous network positions in the process of preferential attachment. Furthermore, the core tags were found to, at different levels, contribute to product sales significantly more than the peripheral tags. The findings support that expertise does not simply reside in individuals but may be socially and dynamically constructed in practice and by communication in larger social units.

Authors

• Larry Zhiming Xu, Annenberg School for Communication and Journalism, U of Southern California

• Matthew Sargent, RAND Corporation

• Yu Xu, U of Southern California

• Jingyi Sun, U of Southern California

• Yiqi (Iona) Li, U of Southern California - Annenberg School for Communication

• Peter Monge, U of Southern California

• Janet Fulk, U of Southern California