A catalogue holds information about a set of objects, typically classified using terms taken from a given thesaurus, and described with the help of a set of attributes. Matching a pair of catalogues means to find a relationship between the terms of their thesauri and a relationship between their attributes. This paper first introduces a matching approach, based on the notion of similarity, that applies to both thesauri and attribute matching. It then describes matchings based on mutual information and introduces variations that explore certain heuristics. Finally, it discusses experimental results that evaluate the precision of the matchings and that measure the influence of the heuristics.

No abstract

Resumo: Evasão de estudantesé um dos principais problemas em cursos de educação a distância. Um dos desafiosé desenvolver métodos para prever o comportamento de estudantes, de forma que professores e tutores possam identificar estudantes em risco de evasão tão cedo quanto possível e prover assistência antes que a evasão ocorra ou o estudante reprove. Modelos de Aprendizado de Máquina tem sido utilizado para prever e classificar estudantes nesses cenários. No entanto, enquanto estes modelos mostram resultados promissores em alguns casos, usualmente utilizam atributos que torna difícil a transferência para outros cursos e plataformas. Neste artigo, provemos uma metodologia para classificar estudantes utilizando apenas contagem de interações de cada estudante. Avaliamos esta metodologia utilizando um conjunto de dados de dois cursos baseados na plataforma Moodle. Executamos experimentos que consistem de treinar e avaliar três modelos de aprendizado de máquina (Máquina de Vetor de Suporte, Bayes Ingênuo e Adaboost comárvores de decisão) em diferentes cenários. Provemos evidências que padrões contidos nas interações podem prover informaçõesúteis para classificar estudantes em risco. Esta classificação permite a personalização de atividades apresentadas a estes estudantes (de forma automática ou através de tutores) como forma de tentar evitar a evasão. Palavras Aprendendo a Identificar Estudantes em Risco em Educação a Distância Usando Contagem de InteraçõesAbstract: Student dropout is one of the main problems faced by distance learning courses. One of the major challenges for researchers is to develop methods to predict the behavior of students so that teachers and tutors are able to identify at-risk students as early as possible and provide assistance before they drop out or fail in their courses. Machine Learning models have been used to predict or classify students in these settings. However, while these models have shown promising results in several settings, they usually attain these results using attributes that are not immediately transferable to other courses or platforms. In this paper, we provide a methodology to classify students using only interaction counts from each student. We evaluate this methodology on a data set from two majors based on the Moodle platform. We run experiments consisting of training and evaluating three machine learning models (Support Vector Machines, Naive Bayes and Adaboost decision trees) under different scenarios. We provide evidences that patterns from interaction counts can provide useful information for classifying at-risk students. This classification allows the customization of the activities presented to at-risk students (automatically or through tutors) as an attempt to avoid students drop out.

Gazetteers are catalogs of geographic features, typically classified using a feature type thesaurus. Integrating gazetteers is an issue that requires some strategy to deal with multiple thesauri, which represent different classifications for the geographic domain. This paper proposes an instancebased approach to define mapping rates between terms of distinct feature type thesauri in order to enable the reclassification of the data migrated from one gazetteer to another.

The economic effects of isolation policies resulting from the COVID-19 pandemic have led small and medium-sized enterprises (SMEs) to look for alternatives to survive. Within this crisis scenario, an engaged university has an important role to play in a regional context in addressing not only health issues, but also any resultant social and economic problems. An engaged university needs to take actions that go beyond its traditional missions of education and research - it has to deliver knowledge to society. This paper analyzes a university-community project in Brazil to identify the necessary elements that help promote a regionally-engaged university: the SOS-PME Advisory Network project, which was originally designed to assist SMEs during the crisis. As a result, we identified elements necessary for promoting the university’s third mission - social engagement by way of a university-community project: an engaged team, multidisciplinarity, project management, agility, alliances, a communication strategy, institutional support, and reputation.