Alan Ramírez-Noriega

Evaluation module based on Bayesian networks to Intelligent Tutoring Systems

Abstract Assessing knowledge acquisition by the student is the primary task of an Intelligent Tutoring System (ITS). Assessment is needed to adapt learning materials and activities to students capacities. In this paper, a proposal to infer the level of knowledge possessed by the student is presented. A general structure of an ITS is shown, an evaluation module based on Bayesian network is proposed. The module mainly based on a test was implemented to know what student knows. During the test, the software system chooses the new questions based on the responses to the previous ones, that is, the software system makes an adaption in real time. A network of concepts was used to get the inferences, which contains the relationships between concepts. Evaluation module could infer many questions and concepts through the relations and the probabilistic inference of the Bayesian network. It information easily can be used to reinforce weak topics in order to cover the students needs. Given the positive evidence is considered that testing the rest of variable examined in the Bayesian network can provide better accurate in the diagnostic of student’ knowledge possession.

An Affective Learning Ontology for Educational Systems

In recent years, education is characterized by promoting independent study through computational tools, such as Intelligent Tutoring Systems. It is important these educational systems consider affective aspects because it improves the learning process. However, the literature reports that there is no an affective model to guide the design of affective educational systems. For that reason, this work introduces a novel affective learning ontology that examines student, tutor and dialogue issues. Also, this paper presents an assessment of which elements of our ontology are taken into account in education systems such as Intelligent Tutoring Systems. We believe the findings from this research will serve as a useful reference for the design of educational systems such as Intelligent Tutoring Systems.

A feedback system to provide affective support to students

Learning support systems need to provide direct and affective interaction to students; it can encourage them and improve their motivation to learn. However, several systems do not include this kind of interaction between students and tutors. In those cases, the student learning interest and motivation to learn could be negatively affected, and student profits from the system could be impoverished. Only a few works have examined the influence of written affective feedback, but these studies focused on student learning outcomes, not in student motivation to learn. The objective of this work is to analyze the impact of an affective feedback system on the students’ perception of two aspects: (i) the degree to which study time is enjoyable and (ii) the motivation to learn. Results show that the affective feedback system makes more enjoyable study time and motivates students to learn, particularly to students with low academic performance. These system enhancements eventually would improve student‐learning outcomes.

Estimating User Stories’ Complexity and Importance in Scrum with Bayesian Networks

Planning Poker is a light-weight technique for estimating the size of user stories, in face-to-face interaction and discussions. Planning Poker is generally used with Scrum. Planning Poker has a lot of benefits, however, this method is not entirely efficient because the result is always based on the observation of an expert. This paper proposes a new model to estimate the complexity and importance of user stories based on Planning Poker in the context of Scrum. The goal of this work is to facilitate the decision-making of newbie developers when they estimate user stories’ parameters. Hence, the decision of each member would be clearer to understand than when the complexity is taken as a whole. We use a Bayesian Network to co-relate factors to have accurate in the estimation. The Bayesian Network gives the complexity of a user story, according to the Fibonacci scale used in Planning Poker.

User and Context Information in Context-Aware Recommender Systems: A Systematic Literature Review

Using contextual information inside recommendation systems is an effective approach to generate more accurate recommendations. This paper present a review conducted to identify what user’s and context’s information it’s considered relevant by researchers to generate contextual recommendations from 2012 to 2015, based on Kitchenham systematic literature review methodology. The results indicated that there is a large set of possible user’s and context’s information that can be used to do recommendations. This review can be taken as basis for future context-aware recommender systems development, as well as development of contextual user models.

Using Bayesian Networks to Obtain the Task's Parameters for Schedule Planning in Scrum

Planning Poker is a consensus-based technique mostly used for estimating effort or relative size of software development goals. This is applicable to estimate the size ofuser stories, developing releases and iteration plans. It is used generally with Scrum. Planning Poker has a lot of benefits, however, this method is not entirely efficient because the result is always based on the observation of an expert. This paper proposes the identification and validation of the most important factors taken into account by Scrum teams to assign complexity and importance to tasks, which are two important factors to consider for task scheduling. We define a knowledge structure that represents different aspects and the influence of each factor in the final decision. We use a Bayesian Network to co-relate these factors in order to have accurate in the estimation. The Bayesian Network give us the complexity of a task according to Fibonacci scale. Using our technique, software teams can focus in the most important tasks to obtain a quality software.

Knowledge Representation in Intelligent Tutoring System

This study presents a compilation of techniques for Knowledge Representation (KR) in Intelligent Tutoring System (ITS). Shows pros and cons of each approach in order to use the proper technique according to the needs. Analyses literature related to ITS and KR to find the approaches. Highlights: Fuzzy Cognitive Maps, Bayesian Network, Semantic Networks, Graphs, among other methods. Each approach contributes with elements to model knowledge. We made a comparison of each model with determined factors. Each technique of KR provides his own vision of how the world should look. Besides, it shows what information is necessary to represent and what is important to ignore. Different approaches to intelligent reasoning lead to different goals and definitions of success.

Integrating affective learning into intelligent tutoring systems

Affectivity has influence in learning face-to-face environments and improves some aspects in students, such as motivation. For that reason, it is important to integrate affectivity elements into virtual environments. We propose a conceptual model that suggests which elements of tutor, student and dialogue should be integrated and implemented into learning systems. We design an ontology guided by methontology, and apply a mathematical evaluation (OntoQA) to determine the richness of the proposed model. The mathematical evaluation states that the proposed model has relationship richness and horizontal nature. We developed a software application implementing the conceptual model in order to prove its effectivity to generate students’ motivation. The findings suggest that the implemented affective learning ontology impacts positively the motivation in students with low academic performance, in female students and in engineering students.

An Empirical Study on Usability Operations for Autistic Children

Leo Kanner defined autism as a distinct personality disorder, a great difficulty to communicate with others. Some autistic persons show cognitive and motor difficulties. Due to the problem presented in users with autism, it is necessary to create interfaces and applications that fit the capabilities of this type of user. This paper identifies operations to perform tasks in mobile applications that an autistic child can execute with less complexity based on the metrics KLM-GOMS and TLM-GOMS. Mobile applications were tested to calculate such metrics in different use cases. The results presented in this paper show how autistics users from different autism levels perform different type of tasks; the results are compared with the model proposed by GOMS. The experimentation shows the total time in seconds that it is necessary to realize every use case and be able to complete the goal assigned for each application. Our results show that an autistic user need more time to interact with the technology; this address to adjust the classification of the operations for interacting with computer-based systems and develop applications that they adapt to the autistic user.

Affectivity Level for Intelligent Tutoring System Based on Student Stereotype

Affectivity is an important part of the learning process. Several authors integrate affectivity in ITS. However, they provided the same affectivity for all students. Authors realized that students were not benefit in the same way by affectivity. This could be because these systems provided the same affectivity to all students, and every student is different and has different needs. For that reason we propose to create a student stereotype to assign an affectivity level considering gender, academic performance and personality. Based on a questionnaire survey applied to 243 students we concluded that female irregular students are the students who need more affectivity and male regular extrovert students need moderate affectivity.