Applying NLP Metrics to Students’ Self-Explanations

Applying NLP Metrics to Students’ Self-Explanations

G. Tanner Jackson (Arizona State University, USA) and Danielle S. McNamara (Arizona State University, USA)
DOI: 10.4018/978-1-60960-741-8.ch015


Intelligent Tutoring Systems (ITSs) are becoming an increasingly common method for students to engage with and learn course material. ITSs are designed to provide students with one-on-one learning that is tailored to their own pace and needs. These systems can adapt to each users’ individual knowledge and ability level to provide the most pedagogically effective learning environment. Tutoring systems have been designed that cover a variety of topics, including both well-defined and ill-defined domains. ITSs have seen great success within well-defined domains, where the topic itself provides only a limited set of responses. For example, in the domain of algebra, there is a limited set of possible actions that can be performed to solve for an unknown variable. Knowing this complete set of actions allows the tutoring system to predict all possible responses from the user. In contrast, ill-defined domains are more abstract and open ended. Reading comprehension is an ill-defined, open ended domain that can incorporate text from any subject, and involve numerous processes and problems for the learner. The number of associations that learners can make with a given text (e.g., based on personal memories, previous courses, ideas within different parts of the same text, etc.) is virtually infinite. These associations make it almost impossible to predict how a user will respond to a text. In addition to working with more abstract concepts, ITSs within ill-defined domains often have the added challenge of interpreting natural language user input. Incorporating natural language allows learners to use their own words and ideas as they interact with the content; however, this also increases the ambiguity of the interaction and decreases the system’s ability to build a precise model of the learner. Building an accurate learner model is essential for the system to adapt the interaction in a pedagogically appropriate manner.
Chapter Preview


iSTART is a web-based tutoring system designed to improve high school and college students’ reading comprehension by providing instruction and practice using self-explanation and reading strategies. The iSTART system was originally modeled after a human-based intervention called Self-Explanation Reading Training, or SERT (McNamara, 2004; McNamara & Scott, 1999; O’Reilly, Best, & McNamara, 2004). The automated iSTART system has consistently produced gains equivalent to the human-based SERT program (Magliano et al., 2004; O’Reilly, Sinclair, & McNamara, 2004; O’Reilly, Best, & McNamara, 2004). Unlike SERT, iSTART is web-based, and can potentially provide training to schools or individuals with internet access. Furthermore, because it is automated, it can work with students on an individual level and provide self-paced instruction. iSTART also maintains a record of student performance and can use this information to adapt its feedback and instruction for each student. Lastly, iSTART combines pedagogical agents and automated linguistic analysis to engage the student in an interactive dialogue and create an active learning environment (e.g., Bransford, Brown, & Cocking, 2000; Graesser, Hu, & Person, 2001; Graesser, Hu, & McNamara, 2005; Louwerse, Graesser, & Olney, 2002).

Complete Chapter List

Search this Book: