Ecological and social software to support human learning

点击翻译按钮获取中文摘要

Fragmentation increasingly characterizes computer systems that support human learning. Social fragmentation occurs when learners are divided into disparate virtual learning communities. Technological fragmentation occurs when the learning technology is no longer a monolithic single system but consists of a loose collection of learning objects (the "content") and processes (called agents) gathered "just in time" to help learner(s) achieve their learning goals. With my graduate students and colleagues, I have been investigating a number of issues that arise when building fragmented learning technologies. One issue is allowing fragmented systems to adapt to individual differences among learners. In a fragmented system adaptivity to learners must be highly reactive to context and purpose, thus necessitating a new active approach to learner modelling. A second issue is to discover how agents can model each other and negotiate with each other to achieve various goals. A third issue is the social and information environment that surrounds learners. My past research has investigated characteristics affecting effective social and group interaction. My proposed research investigates two approaches to supporting learners in this environment. In one, the ecological approach, learner models and learner interaction data are automatically associated with the learning objects in the environment and then "data mined" for patterns that can help achieve various pedagogical purposes (such as finding individualized content for learners). In the other, a social computing approach, tools are provided for learners to annotate the learning objects themselves, and to share these annotations to facilitate their learning. These issues are being explored by building working prototypes and trying them with real learners. Research into these issues draws from, and can contribute to, many disciplines: artificial intelligence in education, user modelling and adaptive systems, data mining, artificial intelligence, multi-agent systems, the semantic web, and areas of social science. The ultimate goal is learning technology that can knowledgeably and flexibly support learners, both individually and in groups.