SocraticGraphs: a Free-form Interactive Graphical Recognition System

该项目是建立在先前的工作（DUE-0735655），产生OrganicPad，一个自由形式，互动，基于平板电脑的化学结构绘图程序。PI正在开发SocraticGraphs --一个基于网络的跨平台系统，该系统识别学生自由形式的输入，并为学习者提供分层的上下文、“苏格拉底式”反馈和支持，因为他们构建图形表示，而不受基于计算机的评估和辅导系统的限制。SocraticGraphs提供了一个独特的基于图形的交互式学习系统，旨在i）以自然主义的方式识别和响应学生的行为，ii）促进学生如何理解科学概念的研究，开发分析表示，重新表达和解释数据。该系统适用于生物学、化学、物理学、地球科学、数学和工程学等广泛的学科。每项活动的设计都以初步研究为指导，使用铅笔和纸张教程，学生访谈和学生通过在线材料的初步版本工作时的直接观察。PI正在开发一系列化学，物理和分子生物学方面的活动，每一项活动都将通过上下文苏格拉底反馈进行探索和回应。学生的反应数据被记录并存储在数据库中，用于使用包括隐马尔可夫模型和人工神经网络在内的许多数据挖掘技术进行分析。最终，该项目提供的活动是对学生“即时”的自由形式投入作出反应，目的是发展分析思维技能和对科学现象的有力和严格的理解。

This project is building on prior work (DUE-0735655) that produced OrganicPad, a free form, interactive, tablet-PC-based chemical structure drawing program. The PIs are developing SocraticGraphs - a web-based, cross platform system that recognizes student free-form input and provides tiered levels of contextual, "Socratic" feedback and support for learners as they construct graphical representations without the constraints usually imposed by computer based assessment and tutorial systems. SocraticGraphs provides a unique graphics-based interactive learning system designed to i) recognize and respond to student actions in a naturalistic manner and ii) facilitate studies on how students come to understand of scientific concepts, develop analytical representations and re-express and interpret data. This system is applicable to a wide range of disciplines including biology, chemistry, physics, geosciences, mathematics, and engineering. The design of each activity is guided by preliminary research using pencil and paper tutorials, student interviews and direct observations of students as they work through preliminary versions of the on-line materials. The PIs are developing a range of activities in chemistry, physics and molecular biology, each of which will probe and respond with contextual Socratic feedback. The data from student responses are recorded and stored in a database for analysis using a number of data-mining techniques including Hidden Markov Modeling and Artificial Neural Nets. Ultimately, the project is providing activities that respond to student free-form input "on the fly" with the aim of developing analytical thinking skills and a robust and rigorous understanding of scientific phenomena.