Supernetwork-based Methods For Identifying Valuable Users In Enterprise Collaborative Innovation Community

As an important information exchange platform,enterprise collaborative innovation community provides an inexhaustible source of innovation inspiration for enterprises,and an important reference of purchase decision for customers.However,with the rapid development of the Internet community,the frequent interaction between users and the optional expression in generated content exacerbate the information overload and information distortion in enterprise collaborative innovation community,which brings more chances and challenges to user identification and knowledge analysis in online community.To this end,scholars have conducted a series of explorations on methods of user identification and knowledge analysis.However,the existing literatures focus on the use of statistical analysis or social network analysis to study the single information in the community.There is still a lack of heterogeneity framework for expressing and analyzing the multidimensional elements in online community,especially for knowledge elements.At the same time,the existing results for identifying valuable users in the collaborative innovation community,especially for knowledge-based valuable users,have some key disadvantaged such as following: targetmissing results,fuzzy key attributes and single metric for all valuable users.Therefore,this paper proposes a User Knowledge Super-Network Model(UKSNM)based on super-network theory under the context of enterprise collaborative innovation community,to realize the heterogeneous expression of the multidimensional elements,and provide an effective analysis framework for subsequent user identification and knowledge discovery.Furthermore,combined with topic recognition,cluster analysis,genetic algorithm and other data-mining methods,this research analyzes the multidimensional characteristics of users in enterprise collaborative innovation community,and construct a set of efficient methods for valuable user identification and analysis.This paper is mainly including four research topics:First,the construction of User Knowledge Super-Network Model(UKSNM).This study expresses and organizes the user's knowledge at the semantic level by employing topic discovery for user-generated content in community.Furthermore,a User Knowledge SuperNetwork Model(UKSNM)is constructed by integrating the key elements and their mapping relationships in enterprise collaborative innovation community.The proposed model realizes the expression of full-featured elements of a specific knowledge system and provides a good analysis framework for knowledge management.Second,the identification and analysis method of domain experts.Based on the User Knowledge Super-Network Model(UKSNM)and its topology features,the characteristics of knowledge quality and innovation potential for each user in enterprise collaborative innovation community is firstly measured.Furthermore,to find a specific knowledge domain for expert identification,this study exhausts the core knowledge topic and hotspot knowledge topic using UKSNM.And then,this study proposed to combine the index method with UKSNM to identify and analyze the domain expert who adapt to the knowledge innovation task(i.e.design task or technical task)in a specific knowledge filed.Third,the identification method of angel users.Based on the topology features of the User Knowledge Super-Network Model(UKSNM)and the definition of the angel user,this study analyzes and measures the characteristics of knowledge quality,willingness to cooperate and communication ability for each user in enterprise collaborative innovation community.Then clustering methods are employed to discover and analyze the community group,and the angel users are defined using clustering results.Fourth,the identification method of collaborative innovation users.To capture the changes of user behaviors and knowledge interests in the enterprise collaborative innovation community,a Dynamic User Knowledge Super Network Model(DUKSNM)was constructed.Based on DUKSNM,user's knowledge characteristics such as knowledge quality,innovation potential,willingness to cooperate,knowledge complementarity and knowledge collaboration ability are exhausted and measured.Then the multi-objective decision-making model for identifying collaborative innovation users is constructed by including the individual attributes and collaborative attributes of users simultaneously.And genetic algorithm is considered to solve the model.Based on proposed model,this study also introduces a method to identify collaborative innovation users in a specific knowledge domain,for example,hotspot knowledge domain.In summary,this study constructs a super-network model which is more suitable for expressing and analyzing the heterogeneity elements in enterprise collaborative innovation communities.Through the analysis of key elements such as users and knowledge,the methods for identifying and analyzing knowledge-based users such as domain experts,angel users and collaborative innovation users in online community are deeply explored.The validity and usability of our proposed research framework is verified by real data.In theoretical,this thesis expands the method of user identification and knowledge discovery,and also facilitate the application of Super-Network theory in the field of knowledge management.In practice,this research introduces an effective tool for enterprises to obtain valuable users and their knowledge from enterprise collaborative innovation communities,which further provides a good guidance for them to improve the product and manage the community.