Knowledge Discovering And Analyzing Methods On User Innovation Communities Based On Model Construction Of Multiple Knowledge Networks

In the background of rapid development of internet, users with similar interests in improving products or developing innovations in a specific area join together to some innovation communities where they can share what they have developed or innovated to the community, or conduct innovation projects collaboratively. Furthermore, companies are becoming aware of utilizing internet user communities to encourage user innovations. And now a number of scholars have focus on how to encourage the user innovation in online communities. Therefore, the online communities are becoming a promising source of innovation knowledge. User innovation knowledge represents the user needs and inflects the innovation trend of products which is of significance to companies. However, few researches are conducted to study the user innovation knowledge. Therefore, how to identify and utilize the valuable user innovation knowledge is becoming a significant and urgent research issue for companies and researchers.The purpose of this thesis is to discuss and explore the approach of identifying the user innovation knowledge from online user innovation communities based on the weighted knowledge network(WKN) theory. On the basis of the existing studies, a series of research works are conducted and the contributions are as follows:First, with regard to the characteristic of user innovation activities, the thesis introduces the user attention indicator to the current WKN model and thus constructs the bi-weighted user innovation knowledge network(BWKN model). This BWKN model contains two types of node’s weight which are keyword frequency and user attention degree. Keyword frequency represents the supplies of innovation knowledge while user attention degree represents the needs of innovation. Then based on the BWKN model, the core user innovation knowledge and the relations between knowledge are identified by analyzing the nodes, edges, weights and the ego networks.Second, considering the importance of user-experts knowledge to the product and service innovation, this paper proposes a discovering and analyzing method to study the knowledge of user-experts in enterprises’ online user innovation communities based on weighted knowledge network(WKN). First, the knowledge of user-experts, presented as forum posts in virtual communities, was acquired through the method of web content mining, and the acquired knowledge was processed and integrated into a weighted knowledge network model(WKN). Next, based on the WKN model, the basic knowledge mode, the developing knowledge mode and the core knowledge mode of user-experts were identified by analyzing the weights of nodes and edges; the cliques and components in this network. Knowledge mode which is formed when knowledge has frequent co-occurrence reflects the stable relations between knowledge. Compared with the existing method of user knowledge discovering, the presented methods are more clearly and thoroughly to identify the knowledge structure of user-experts, and it also provides a new tool to study the knowledge of user-experts in enterprise virtual communities.Third, besides analyzing the nodes and the relation modes of knowledge, this thesis also presents a new method to discover, model and analyze user-innovation knowledge based on weighted clustered network(WCN). First, character words are mined from the community posts, then are clustered based on the Word-Posts Matrix. Second the WCN model was established by identifying the knowledge nodes(high-frequency keywords nodes and clustered nodes), the weighted value of nodes(frequency of keywords) and the links of nodes(affiliated relations of the nodes). Furthermore, the WCN model is used to analyze the structure of user-innovation knowledge deeply and quantitatively, such as the identification of knowledge clusters, the main sub-fields of user innovation, and the core knowledge clusters of the main sub-fields. The analyzed results can also be illustrated as sub-WCN network models.Fourth, according to the features of innovation activities in online communities, innovations are presented as forum posts by users. The whole innovation involves users, posts, knowledge and thus constitutes the user innovation knowledge system. The knowledge system contains four essences: user, innovation post, innovation knowledge, and innovation knowledge field. Therefore, this thesis proposes the model of user innovation knowledge super-network(UIKSN) to analyze the mapping relations between each essence and then identify the valuable users, posts and knowledge.Fifth, lead users are regarded as the most valuable users according to the user innovation theory. In the background of the rapid development of internet especially the appearance of user innovation communities, lead users could be also identified from these communities. This thesis defines lead users in online communities of new features: knowledge expertise and user attention degree. Here, knowledge expertise means that lead users should be experts in certain innovation field to some degree. User attention degree means that the innovation receives high attentions or compliments from other users. Therefore, based on the former UIKSN model, this thesis proposes a new super-network model(UHKSN) through introducing the user attention degree to the UIKSN. By analyzing the knowledge contributions and user attention degree, lead users who with high user attention degree and high knowledge contributions are identified based on the UHKSN.The proposed methods of model construction and analysis can solve many practical problems, such as the fragmented knowledge integration in online user communities and the identification of lead users. It also improves the constructing and analyzing methods of WKN and WSN model.