Abstract
This dissertation aims to enable the utilization of extensive textual datasets collected through large-scale digital participation in urban transformation contexts. It predominantly focuses on online debates, recognizing that these voluminous datasets are repositories of crucial insights into urban issues from the residents’ perspectives. Despite their potential, current literature and practices in digital participation often underutilize these datasets, primarily due to the challenges in extracting pertinent information and the lack of effective methodologies for qualitative data representation and visualization in urban design and planning.
To surmount these challenges, the dissertation enhances our understanding of ’postparticipation’, underscoring the crucial phase that bridges data collection and its practical application. It advocates for a structured approach to managing complex textual datasets while devising an integrated data analysis workflow. Pursuing this goal, the research meticulously examines participation datasets from three diverse case studies, Singapore, Madrid, and Hamburg, which were selected for their variation in scale (ranging from local to city-wide), initiative type (government versus project-initiated), and governance systems (top-down versus bottom-up). This diversity ensures that the
findings are applicable and generalizable to a broad spectrum of citizen participation processes. Consequently, this dissertation adopts a mixed-methods approach to:
• Define and explore the ’post-participation’ phase;
• Develop methodologies for processing textual datasets;
• Formulate methods for analyzing digital participation data;
• Create an integrated workflow to visualize and analyze large-scale textual data.
This dissertation is grounded in a multidisciplinary approach, integrating state-ofthe-art AI-driven analytical techniques such as Natural Language Processing (NLP),topic modeling, sentiment analysis, and content analysis for effective navigation of textual datasets. Complementing these techniques, the study also incorporates operations research methods, notably Multi-Criteria Decision Analysis (MCDA), to systematically evaluate participation data and refine data collection processes. Building on this methodological foundation, the dissertation presents the investigated textual datasets and their corresponding analysis results through visual representations, enhancing their accessibility and utility. To ensure the relevance and applicability of these methods, the research process includes a rigorous evaluation phase. This phase involves conducting interviews with both experts and residents, aligning the study’s outcomes with the practical needs and concerns of urban practitioners, particularly in the context of post-participation in urban transformation processes.
Backed by these insights, the dissertation enables data-driven decision-making, seamlessly integrating advanced text modeling, comprehensive quantitative and qualitative assessment frameworks, and multi-layered visualization techniques into digital participation. In a practical aspect, it also delves into user interface design and instructional modalities, aiming to democratize the accessibility and utilization of such complex methodologies and workflows among diverse stakeholders. This effort culminates in the development of the Digital Participation Data Analysis Toolkit (DIP-DAT), specifically designed for the post-participation phases. Overall, this dissertation significantly contributes to the field of urban design and planning support through a meta-analytic approach, providing robust analytical tools and methodologies geared towards fostering more sustainable, human-centered, and inclusive built environments.