Network traits driving knowledge evolution in open collaboration systems

Abstract

Network interpretation illuminates our understanding of the dynamic nature of cultural evolution. Guided by cultural evolution theory, this article explores how people collectively develop knowledge through knowledge collaboration network traits. Using network data from 910 artifacts (the WikiProject Aquarium Fishes articles) over 163 weeks, two studies were designed to understand how collaboration network traits drive population and artifact-level knowledge evolution. The first study examines the selection pressure imposed by10 network traits (against 11 content traits) on population-level evolutionary outcomes. While network traits are vital in identifying natural selection pressure, intriguingly, no significant difference was found between network traits and content traits, challenging a recent theory on network-driven evolution. The second study utilizes time series analysis to reveal that three network traits (embeddedness, connectivity, and redundancy) at a prior time predict future artifact development trajectory. This implies that people collectively explore various positions in a potential solution space, suggesting content exploration as a possible explanation of knowledge evolution. In summary, understanding the interplay between network traits and content exploration provides valuable insights into the mechanisms driving knowledge evolution and offers new avenues for future research.