Abstract:Purpose Social loafing in knowledge contribution (namely, knowledge contribution loafing [KCL]) usually happens in group context, especially in the mobile collaboration tasks. KCL shows dynamic features over time. However, most previous studies are based on static assumption, that is, KCL will not change over time. This paper aims to reveal the dynamics of KCL in mobile collaboration and analyze how network centrality influences KCL states considering the current loafing state. Design/methodology/approach This study is based on empirical study design. Real mobile collaboration behavioral data related to knowledge contribution were collected to investigate the dynamic relationship between network centrality and KCL. In total, 4,127 chat contents were collected through Slack (a mobile collaboration APP). The text data were first analyzed using the text analysis method and then analyzed by a machine learning method called hidden Markov model. Findings First, the results reveal the inner structure of KCL, showing that it has three states (low, medium and high). Second, it is found that network centrality positively influences individuals involved in medium and high loafing state, while it has a negative influence on individuals with low loafing state. Research limitations/implications The limitations are related to the single machine learning method and no subdivision of social network. First, this paper only uses one kind of text classification model (TF-IDF) to divide chat contents, which may not be superior to other classification models. This paper considers the eigenvector centrality, and not further divides the social network into advice network and expressive network. Practical implications This study helps companies infer tendency of different KCL and dynamically re-organize a mobile collaborative team for better knowledge contribution. Originality/value First, previous studies based on static assumptions regarding KCL as static and the relationship between loafing reducing mechanisms and team members KCL does not change over time. This study relaxes static assumptions and allows KCL to change during the process of collaboration. Second, this study allows the impact of network centrality to be different when members are in different KCL states.

Leave or Not Leave? Group Members' Departure Prediction in Dynamic Information Networks.

Understanding the Determinants and Dynamic Process of User Exodus in Social Networking Sites: Evidence from Kaixin001

A Dynamic Day-To-Day Departure Time and Route Choice Model for Bounded-Rational Individuals

Predicting group stability in online social networks

Location Prediction: Communities Speak Louder than Friends

Should I Stay or Should I Go: Predicting Changes in Cluster Membership

A Novel User Preference Prediction Model Based on Local User Interaction Network Topology

Come-and-Go Patterns of Group Evolution: A Dynamic Model.

Intelligent Link Prediction Management Based on Community Discovery and User Behavior Preference in Online Social Networks

Dynamic-group-aware networks for multi-agent trajectory prediction with relational reasoning

Promotion and resignation in employee networks

Come N Go

Social network user influence sense-making and dynamics prediction

Group Behavior Prediction and Evolution in Social Networks

Dynamic impact of social network on knowledge contribution loafing in mobile collaboration: a hidden Markov model

Predicting personalized grouping and consumption: A collaborative evolution model

Nonparametric Multi-group Membership Model for Dynamic Networks

Predicting Small Group Accretion in Social Networks: A topology based incremental approach

A Dynamic Logistic Regression for Network Link Prediction

Friendship prediction model based on factor graphs integrating geographical location

Meta-path aware dynamic graph learning for friend recommendation with user mobility