Combinatorial creativity: Knowledge graphs and idea generation in crowdsourcing innovation

• Main Author: MACK, Zhi Wei Vincent
• Format: text
• Language: English
• Published: Institutional Knowledge at Singapore Management University 2024
• Subjects: Knowledge Representation; Knowledge Graph; Idea Semantics; Crowdsourcing Innovation; Idea Generation; Combinatorial Creativity; Artificial Intelligence and Robotics; Graphics and Human Computer Interfaces
• Online Access: https://ink.library.smu.edu.sg/etd_coll/615; https://ink.library.smu.edu.sg/context/etd_coll/article/1613/viewcontent/GPIS_AY2019_PhD_Mack_Zhi_Wei_Vincent.pdf
• Institution: Singapore Management University

This dissertation explores the dynamic interplay between combinatorial creativity and technology-driven innovation within various knowledge-intensive fields. It critically examines the role of combinatorial creativity in generating groundbreaking innovations by amalgamating existing ideas and technologies. This research incorporates a detailed examination of how knowledge, whether tacit or explicit, can be transformed into actionable data to foster innovation in crowdsourcing contexts. Chapter 2 provides an overview of the relevant literature on how Artificial Intelligence and Knowledge Management Systems can support combinatorial creativity. The study further delves into the transformative impact of knowledge management systems, particularly focusing on crowdsourcing platforms that leverage collective intelligence to accelerate the innovation cycle. Chapter 3 examines how unstruc- tured text can be structured into knowledge graphs using Conceptual Dependency (CD) theory (Schank 1969, 1975) to create explicit knowledge representations from tacit knowledge. A natural language parser was developed that identifies and extracts conceptual relationships to form ”conceptual molecules”. Subsequently, the extracted conceptual relationships were organized algorithmically into knowledge graph triples. This knowledge graph construction approach was applied successfully within the context of TVTropes.org, and the resultant output of conceptually meaningful triples were integrated into a graph database. By designing specific graph queries, this chapter demonstrated that the resultant knowledge graph was capable of being used for information retrieval purposes, as well as to recommend knowledge elements for the generative stage of combinatorial creativity. Overall, this chapter demonstrates the viability of a Conceptual Dependency Theory approach to designing unsupervised rule-based knowledge graph construction systems. Building on the structural aspects of knowledge graphs, Chapter 4 explores how the network connections within these graphs facilitate access to existing knowledge, moderated by the content characteristics of the knowledge graph nodes, that can foster the creation and exchange of successful ideas in a crowdsourcing environment. By studying the citation networks on TVTropes.org, particularly within the Trope Launch Pad subwiki, how the structural properties of the trope’s citation network such as embeddedness and bridging can influence idea success (which is defined in Sections 4.4.2.1 and 5.4.2.1) were analyzed. The research found that the semantic diversity of the trope’s existing knowledge base significantly moderates these effects, which contributes to the success of ideas in a crowdsourcing environment. Furthermore, the direct and combined effects of the structural and semantic properties of the knowledge base are mostly consolidated through the idea refinement phase in the crowdsourcing process, ultimately impacting their success within the crowdsourcing ecosystem. Finally, Chapter 5 brought the focus to the interactions between ideators and evaluators during the refinement phase within crowdsourcing platforms. Through an examination of discussion forums, nuanced dynamics were unraveled, revealing how the active reactions of ideators to feedback influences both idea success and community engagement. Findings reveal that while ideators’ responses generally boost community engagement, their impact on idea success is moderated by the semantic proximity of the feedback. Distant feed- back encourages innovation, whereas closely aligned feedback tends to enhance community engagement more effectively. On the other hand, while feedback integration is generally associated with both idea success and community engagement, integrating distant feedback tends to result in a negative effect on idea success.