Developing an Intelligent Tutoring System that has Automatically Generated Hints and Summarization for Algebra and Geometry

Developing an Intelligent Tutoring System that has Automatically Generated Hints and Summarization for Algebra and Geometry

Yatao Li, Ke Zhao, Wei Xu
DOI: 10.4018/ijicte.2015040102
(Individual Articles)
No Current Special Offers


Intelligent tutoring systems (ITSs), which provide step-by-step guidance to students in problem-solving activities, have been shown to enhance student learning in a range of domains. However, they tend to be preestablished and cannot supply the tutoring function immediately from the diverse mathematical questions. The MITSAS (multiagent intelligent tutoring system after school) is a web-based ITS in algebra and geometry with a natural language interface which is designed to extract the hint and summarization from the detailed solving answer automatically. In this paper, its Design principles and functionality is analysed firstly. Then, the framework including the natural language understanding agent, automatic modelling agent and automatic problem-solving agent are discussed in the following in order to support the real-time problems solution. Next, the methods for automatically extracting tutoring function such as hint and summarization is given based on the difficulty of knowledge components and the type of problem acquired from the detailed answer. Finally, the effectiveness of MITSAS at improving the Chinese Students' learning gain is shown by an experiment conducted in junior school.
Article Preview


Web-based ITS is gaining popularity with the advances of information technology and artificial intelligence. In ITS, the agent or set of agents can be modeled to perform pedagogical or tutoring tasks. The interactions among these agents and students include instructing, evaluating feedback from students, mining the characteristics of students, solving problems and effectively tutoring with hint. Thus, these agents have to be able to both instruct and tutor at the same time and also have the ability to cope with disparate learning environments adaptively.

As early as 1956, research by Prof. Benjamin Bloom and others demonstrated that students who receive one-on-one instruction perform two standard deviations better than students in traditional classrooms (Bloom, Engelhart, Murst, Hill, Drathwohl, 1965). That is, the average tutored student performed as well as the top 2% of those receiving classroom instruction. Furthermore, research on prototype systems indicates that students taught by ITS generally learn faster and translate the learning into improved performance better than classroom-trained participants. In the 1990's, references (Lynnette Taylor, 1999; Wood, Underwood, Avis, 1999; Underwood, Cavendish, Dowling, Fogelman, Lawson, 1996) conducted a formal evaluation of a two year trial of two Integrated Learning Systems in United Kingdom schools. It was shown that the systems could improve learning performance significantly.

Based on the learner model, ITS tailor instructional strategies, in terms of both content and style, and provide explanations, hints, examples, demonstrations and practice problems as needed. The results that there were high learning gains for both the effectiveness of the human teachers and the computer-based hinting e-learning system even without the use of adaptive and personalization capabilities were given in literature (Muñoz-Merino, Kloos and Muñoz-Organero, 2011; Henk, Harskamp, Suhre, and Goedhart, 2009; Suebnukarn & Haddawy, 2006). Some examples of hinting tutors are Andes (VanLehn et al., 2005), SIETTE (Guzma & Conejo, 2005), PACT(Aleven, Koedinger, and Cross, 1999), or AgentX (Martin, & Arroyo, 2004). In web-based learning environments for mathematics, such as Assistments (Razzaq, Heffernan, Feng, and Pardos, 2007) and ActiveMath (Melis, Andre`s, Bu¨denbender, Frischauf, Goguadze, Libbrecht, Pollet, and Ullrich, 2001), they generally supply richer client functions such as hint than HTML- and Javascript-based interfaces based on Flash’s ActionScript programming capabilities. The Mathtutor (Aleven, McLaren, and Sewall, 2009) offers detailed, interactive, step-by-step guidance with problem solving, individualized problem selection, detailed reports of student performance for teachers, parents, etc. on the basis of cognitive model. Cognitive Tutors provide individualized support for guided learning by doing (Anderson, Corbett, Koedinger & Pelletier, 1995; Koedinger, & Aleven, 2007) with context-sensitive hints and instruction to guide students towards reasonable next steps.

However, most of these ITSs ignore the summarization of the knowledge and methods for solving problem. Both hints and summarization are common way to help student solve problem for human tutor. Summarization is conducive to the students to construct the abstract relation of knowledge.

Complete Article List

Search this Journal:
Volume 20: 1 Issue (2024)
Volume 19: 1 Issue (2023)
Volume 18: 3 Issues (2022)
Volume 17: 4 Issues (2021)
Volume 16: 4 Issues (2020)
Volume 15: 4 Issues (2019)
Volume 14: 4 Issues (2018)
Volume 13: 4 Issues (2017)
Volume 12: 4 Issues (2016)
Volume 11: 4 Issues (2015)
Volume 10: 4 Issues (2014)
Volume 9: 4 Issues (2013)
Volume 8: 4 Issues (2012)
Volume 7: 4 Issues (2011)
Volume 6: 4 Issues (2010)
Volume 5: 4 Issues (2009)
Volume 4: 4 Issues (2008)
Volume 3: 4 Issues (2007)
Volume 2: 4 Issues (2006)
Volume 1: 4 Issues (2005)
View Complete Journal Contents Listing