Multi-label classification of member participation in online innovation communities

Abstract

Online innovation communities are defined as internet-based platforms for communication and exchange among customers interested in building innovations for a given product or technology. As firms recognize an online innovation community as a valuable resource for integrating external consumer knowledge into innovation processes, they increasingly ignore to build long-term interactions and collaborations. However, in the pursuit of a long-term community, moderators face enormous challenges, especially due to inferior member participation. Inferior member participation, whether in the form of inferior participation quantity, quality and/or emotionality, produces a community with minimal activity, unhelpful content and a nonconstructive atmosphere, respectively. Because members can be associated with multiple labels of inferior participation behavior simultaneously, the paradigm of multi-label (ML) classification methodology naturally emerges, which associates each member of interest with a set of labels instead of a single label as known in traditional classification problems. Using 1407 members of 7 real-life innovation communities, this study explores 10 state-of-the-art ML algorithms in an extensive experimental comparison to explore the benefit of ML classification methodology. We advance literature by demonstrating a novel application for ML classification adoption in the domain of online innovation communities, while comparing ML classifiers in the smallest possible scenario of 3 labels. The results indicate the effectiveness of the ML classification methodology for inferior member participation prediction, gives insights into ML classifiers’ performance and discusses paths for future research.