Question Accumulation and Response Tool for Education and Training

Students in distance learning classes have at least as much need to ask questions of their instructors as students in more conventional settings. However, distance instructors are not always available when students need help. Across a course and its repeat offerings, many questions recur frequently. This provides possible leverage for incrementally constructing custom question-answering systems for particular distance learning courses, as a way of making better use of available instructor time. Stottler Henke proposes to apply its extensive experience with intelligent instructional systems, language and knowledge processing, plus its evolving in-house tool-base to build an innovative tool to lower the cost of creating, administering, and maintaining a high precision question-answering system for instructional applications. In this Phase I project, we will (1) explore major technical challenges in areas such as semi-automated learning of questions and answers, and incremental refinement of questions and answers through dialog, (2) design an overall system that uses a diverse set of language processing and machine learning techniques to support storage, retrieval, and construction of accurate question answers, and (3) build a limited proof-of-concept prototype that illustrates the utility of our proposed approach to constructing a QUestion Accumlation and Response Tool for Education and Training (QuARTET).