580 California St., Suite 400
San Francisco, CA, 94104
Key research themes
1. How can computational models and AI techniques optimize the construction and assessment of knowledge spaces in educational contexts?
This research focuses on data-driven and computational methods for constructing, modeling, and assessing knowledge spaces, especially for adaptive learning and student modeling. These approaches aim to optimize the representation of students' knowledge states and enable efficient, accurate assessment and prediction of learner performance. Leveraging machine learning, neuroevolution, and probabilistic graphical models, the studies address challenges such as scalability, complexity, and computational efficiency in building large, realistic knowledge spaces useful in intelligent tutoring and educational assessments.
Key finding: Proposes a novel neuroevolutionary data-driven approach to constructing large knowledge spaces by analogy between neurons in neural networks and knowledge states, showing superior scalability compared to traditional inductive... Read more
Key finding: Provides empirical evidence from over 210,000 assessments that knowledge spaces can comprehensively and effectively represent students' mastery in topics like beginning algebra, with predictions of student responses to unseen... Read more
Key finding: Demonstrates that item-to-item based probabilistic models, specifically POKS (naive Bayes framework) and Bayesian networks, can be accurately learned from data to predict student item responses with high efficiency. POKS... Read more
2. How do social and collaborative knowledge spaces supported by digital environments enhance collective knowledge building and learning?
This theme investigates the design, use, and impact of digital knowledge spaces that facilitate collaborative knowledge construction in educational and organizational settings. Emphasizing socio-cognitive and cultural dynamics, these studies explore how shared digital platforms support distributed learning communities, enable knowledge co-production, and transform traditional pedagogy through social knowledge workspaces. The research highlights affordances such as active knowledge production, metacognitive reflection, and collaborative intelligence enabled by pervasive digital tools.
Key finding: Identifies the Joint Problem Space (JPS) as a shared conceptual structure that emerges through collaborative problem solving mediated by computer simulation. The JPS integrates semantic interpretations of goals and operators,... Read more
Key finding: Establishes the conceptual foundation of knowledge building environments as networked collaborative spaces where communities co-produce, refine, and advance ideas collectively. Empirical findings indicate that these cyber... Read more
Key finding: Presents Scholar as a social knowledge technology designed to enable collaborative intelligence through pedagogical affordances such as ubiquitous learning, active production, and recursive feedback. The work bridges the... Read more
3. What roles do physical and semantic knowledge spaces play in organizing, managing, and disseminating complex knowledge in organizations and urban contexts?
This area addresses the conceptualization, design, and evaluation of knowledge spaces as environments—both physical and semantic—that facilitate knowledge integration, access, management, and innovation. It includes frameworks for enterprise knowledge architectures, semantic web tuplespaces, learning spaces in educational institutions, and urban knowledge hubs. The work emphasizes multidimensionality, interdependence of information, roles, tasks, and systems in knowledge management, and explores how emerging technologies and urban policies shape knowledge ecosystems and knowledge economies.
Key finding: Introduces a four-dimensional framework (information, roles, tasks, views - IRTV) for modeling personal workspaces, innovation, business networking, and community spaces within enterprises. This nested model elucidates how... Read more
Key finding: Analyzes the challenges and progress in establishing semantic knowledge spaces on the Internet, focusing on semantic content availability, ontology development, scalability, and multilingualism. The paper contextualizes... Read more
Key finding: Argues that learning spaces must be reconceptualized as flexible, multi-dimensional environments shaped by and shaping learning communities. Case studies demonstrate how physical configurations, embedded technology, and... Read more