Supportive Automated Feedback for Short Essay Answers (SAFeSEA)

点击翻译按钮获取中文摘要

The aim of this research is to provide an effective automated interactive feedback system that yields an acceptable level of support for university students writing essays in a distance or e-learning context. Our tool will be based on an existing system supporting the online writing and assessment of essays, OpenComment (Whitelock & Watt, 2008), but this requires several problems to be solved, both in natural language processing, and in educational theory and practice. This proposal is therefore a truly interdisciplinary one.The natural language processing problems are how to `understand' a student essay well enough to provide accurate and individually targeted feedback, and how to generate that feedback automatically. We intend to use techniques from document classification (measures of semantic similarity), information extraction, document summarisation, and text generation to address these problems. The tools we build need to be flexible enough to support different types and models of feedback.The educational problem is how to develop and evaluate effective models of feedback. This requires research into the selection of the content, mode of presentation and delivery of the feedback. The primary aspects of feedback which will be investigated are: summarisation; recognition of positive achievements; location of errors of commission or omission; misconceptions or problems connected with causal relationships; tactical and strategic hints both relating to the specific issues being addressed in the essay and issues connected with `metacognition' (i.e. self awareness of the learning process), and self-regulation of the learning process. For feedback to be effective, it should assist students not only to manage their current essay-writing task, but it should also lead to further development of their essay-writing skills, the skills associated with efficient self-regulation, and their motivation to complete their course. The methods used to obtain data on effectiveness will be student surveys, interviews with selected participants, analysis of the various versions of the essays submitted for feedback, and the quality of the final essay.