Learning to Perceive Two-Dimensional Displays Using Probabilistic Grammars

Abstract

Abstract. People learn to read and understand various displays (e.g., tables on webpages and software user interfaces) every day. How do humans learn to process such displays? Can computers be efficiently taught to understand and use such displays? In this paper, we use statistical learning to model how humans learn to perceive visual displays. We extend an existing probabilistic context-free grammar learner to support learning within a two-dimensional space by incorporating spatial and temporal information. Experimental results in both synthetic domains and real world domains show that the proposed learning algorithm is effective in acquiring user interface layout. Furthermore, we evaluate the effectiveness of the proposed algorithm within an intelligent tutoring agent, SimStudent, by integrating the learned display representation into the agent. Experimental results in learning complex problem solving skills in three domains show that the learned display representation is as good as one created by a human expert, in that skill learning using the learned representation is as effective as using a manually created representation.

A New Student Modeling Technique With Convolutional Neural Networks: LearnerPrints

“…Therefore, a student model should be able to accurately represent a student and predict the students’ needs (Chrysafiadi & Virvou, 2013). Student modeling also enables instructional decisions to be made in learning environments (Li et al., 2011). There are different approaches to student modeling.…”

“…For example, the overlay model represents the student’s knowledge level, the stereotype model focuses on the student’s characteristics, the perturbation model concentrates on the student’s knowledge and misconceptions, and machine learning techniques are used for automated observation of students’ actions and behavior (Chrysafiadi & Virvou, 2013). Traditional student modeling approaches are subjective; they are time consuming and require expert opinion (Li et al., 2011). Therefore, the use of machine learning techniques in the student modeling process is becoming widespread.…”

“…They used a deep/surface approach to learning in order to classify students. Li et al. (2011) used machine learning techniques to create a student model automatically and estimate student performance.…”

“…Although the learning process is individual, a student model can be created according to the characteristics of the students or the behavior patterns exhibited in the learning environment. Student modeling is key to ensuring effective teaching in intelligent tutoring systems, adaptive educational hypermedia and personalized learning environments (Chrysafiadi & Virvou, 2013; Kaser et al., 2017; Li et al., 2011). In creating a personalized learning process in these environments, student modeling is the most important component (Güyer & Çebi, 2015; G. Yang et al., 2010).…”

“…Today, machine learning techniques can be easily applied in e-learning environments using tools developed for machine learning such as scikit-learn (Pedregosa et al., 2011), Keras (Chollet, 2018), and Weka (Frank et al., 2016). In educational environments, these techniques are used for predicting student performance (Al-Shehri et al., 2017; Aydoğdu, 2020), student modeling (Li et al., 2011; Webb et al., 2001) and automatic detection of off-task behaviors (Cetintas et al., 2010). Artificial neural networks, which are one group of machine learning techniques, are some of the most frequently used techniques in educational environments (Manjarres et al., 2018).…”