Research On The Influence Mechanism Of Human-bot Collaboration Patterns In Online Knowledge Community

With the development of Web2.0,online knowledge community(OKC)has become a large-scale collaborative platform for users to produce and share knowledge.With the continuous development of artificial intelligence(AI),OKC is gradually changing from humanhuman collaboration to human-bot collaboration for knowledge production.However,although bots in the current community have become one of the important cooperative actors,the research on the influence mechanism of human-bot collaboration is still lack of enough attention.In addition,the current research on human-bot collaboration lacks the corresponding theoretical framework as a guide,and lacks the research on how the different human-bot attributes affect the formation of human-bot collaboration patterns through different types of dependencies.Therefore,this paper studies the influence mechanism of human-bot collaboration patterns in OKC from the perspective of dependencies between resources.Firstly,based on the coordination theory,this paper constructs the research model of human-bot collaboration mechanism in OKC to study the influence relationship between humans' knowledge attribute,bots' skill attribute,different types of dependencies,task type and human-bot collaboration patterns.Secondly,this paper takes the human-bot collaboration log data in Wikipedia as the research sample,uses GRAMI subgraph mining algorithm to mine the dependencies between resources from the knowledge resource network formed by humanbot collaboration,and uses local process mining(LPM)algorithm to mine the frequent humanbot collaboration patterns in the human-bot collaboration process,and the collaboration patterns are classified into different types.Then,this paper uses partial least squares regression analysis and mediating effect test empirical methods to test the research model of human-bot collaboration mechanism in OKC,and finds out the influence mechanism of humans' knowledge attribute and bots' skill attribute on the formation of human-bot collaboration patterns through different types of dependencies.Finally,this paper uses the social computing experimental method to further verify and compare the empirical research conclusions by constructing the human-bot collaboration simulation model in Wikipedia,and reveals the influence mechanism of human-bot collaboration processes in OKC.Based on the results of empirical and computational simulation research,this paper explains the influence of human-bot attributes on dependencies in OKC,and the influence of human-bot attributes on different human-bot cooperation modes through different types of dependencies,verifies and expands the application scenarios of coordination theory in largescale human-bot collaboration environment in OKC,and further enriches the research on human-bot collaboration.Finally,it also provides the corresponding enlightenment for the community managers to manage the human-bot cooperation process in OKC.