Prerequisite identification via knowledge graphs

The identification of prerequisite relationships among concepts is a fundamental step toward the organization of knowledge for educational purposes. The identification of such prerequisite relations is a crucial step for instructional designers and for the next generation of automated systems whose goal is to support the generation, recommendation, and adaptation of learning paths. In this thesis, we address the problem of the automatic identification of the prerequisite relationships among concepts. We explore two different strategies. The first is a simple measure that analyzes links between the concepts in a Knowledge Graph base belonging to the linked open data initiative. The second strategy uses machine learning methods to identify the prerequisite relationship based on a set of features extracted mostly from the Knowledge Graph. The proposed approach is validated on three benchmarks concept pairs datasets in different domains. The results show the superiority of the machine learning method and that it is possible in most cases to automatically discern the prerequisite relationship. Moreover, our approach overcomes other supervised and unsupervised strategies that have been proposed in the literature for this problem.

• Autores: Manrique Piramanrique, Rubén Francisco
• Fecha de publicación: 2019
• Institución: Universidad de los Andes
• Repositorio: Séneca: repositorio Uniandes
• Idioma: eng
• OAI Identifier: oai:repositorio.uniandes.edu.co:1992/44020
• Acceso en línea: http://hdl.handle.net/1992/44020
• Palabra clave: Grafos de conocimiento; Web semántica; Estructuras conceptuales (Teoría de la información); Ingeniería
• Rights: openAccess
• License: http://creativecommons.org/licenses/by-nc-nd/4.0/