Personalization Approaches in Learning Environments
Lecture Notes in Computer Science
Personalization approaches in learning environments can be addressed from different perspectives and also in various educational settings, including formal, informal, workplace, lifelong, mobile, contextualized, and selfregulated learning. PALE workshop offers an opportunity to present and discuss a wide spectrum of issues and solutions. In particular, this fourth edition includes 8 papers dealing with student's performance, modeling the user profile in a standardize way, computing attributes
... r learner modeling, detecting affective states to improve the personalized support, and applying user modeling approaches in new contexts, such as MOOCs and gamified environments. PALE 2014 (Edited Abstract. Traditionally, the assessment and learning science communities rely on different paradigms to model student performance. The assessment community uses Item Response Theory which allows modeling different student abilities and problem difficulties, while the learning science community uses Knowledge Tracing, which captures skill acquisition. These two paradigms are complementary -IRT cannot be used to model student learning, while Knowledge Tracing assumes all students and problems are the same. Recently, two highly related models based on a principled synthesis of IRT and Knowledge Tracing were introduced. However, these two models were evaluated on different data sets, using different evaluation metrics and with different ways of splitting the data into training and testing sets. In this paper we reconcile the models' results by presenting a unified view of the two models, and by evaluating the models under a common evaluation metric. We find that both models are equivalent and only differ in their training procedure. Our results show that the combined IRT and Knowledge Tracing models offer the best of assessment and learning sciences -high prediction accuracy like the IRT model, and the ability to model student learning like Knowledge Tracing. Abstract. In this paper, we focus on user assistance in an interactive and adaptive system. The increase in production of digital data, these last two years has raised several issues regarding the management of heterogeneous and multiple source data in the user's environment. One distinctive feature of user assistance is the user model that represents essential information about each user. We propose a modeling of a scientific user, who is a researcher, in a personal resource management system. Our methodology is based on the IMS-LIP standard extension and the user's trace management. Our work can assist users in the consolidated management of their resources and their environment, based on the user's profile. Experimental results obtained by the emperical evaluation in our laboratory are presented. Abstract. Collaborative indicators derived from quantitative statistical indicators of students' interactions in forums can be used by e-learning systems in order to support the collaborative behaviour and motivation of students. The main objective of this research is to achieve a transferable and domain-independent reputation indicator, considering the information extracted from social network analysis, statistical indicators, and opinions received by students in terms of ratings. This paper describes how to consider the reputation indicator in a collaborative environment in order to group students (distributing the most prominent students into different groups) aimed to improve the collaborative indicators (such as initiative, activity, regularity). Abstract. We report on a project with the goal of creating a proactive system that attempts to reduce the propensity to mind wander (MW) by optimizing learning conditions (e.g., text difficulty and value) for individual learners. Our previous work had shown that supervised classification based on individual attributes could be used to detect the learning condition with the lowest MW rates. Here we test the model by comparing MW rates for the predicted optimal conditions to MW rates from a random control condition or in the condition with the overall best MW rate across all learners. Our results suggest that our method is better than these non-adaptive alternatives in certain contexts. Abstract. Supervising a student's resolution of an arithmetic word problem is a cumbersome task. Different students may use different lines of reasoning to reach the final solution, and the assistance provided should be consistent with the resolution path that the student has in mind. In addition, further learning gains can be achieved if the previous student's background is also considered in the process. In this paper, we outline a relatively simple method to adapt the hints given by an Intelligent Tutoring System to the line of reasoning that the student is currently following. We also outline possible extensions to build a model of the student's most relevant skills, by tracking user's actions. Abstract. Automated detection of constructs associated with student engagement, disengagement, and meta-cognition plays an increasingly prominent part of personalized online education. Often these detectors are trained with ground truth labels obtained from field observations, a method that balances collection speed with label quality. Some behaviors and affective states (e.g., boredom) are regularly modeled across learning environments, but other constructs (e.g., gaming the system) manifest in fewer systems. New environments create the possibility of entirely unexpected constructs. In this paper, we describe how a field observation protocol (already proven effective for affect and behavior detection in several systems) was adapted to provide the flexibility needed to document previously unidentified or rare constructs. Specifically, we describe the in-field modification of the Baker Rodrigo Ocumpaugh Monitoring Protocol (BROMP) to accommodate categories not previously established (e.g., creative metanarrative) during observations of an educational multi-user virtual environment (MUVE). We also discuss the importance of developing methods that allow researchers to conduct such explorations while still capturing standard data constructs.