<abstract xmlns="http://www.w3.org/1999/xhtml">

<p>This paper proposes a personalized teaching method based on knowledge mapping and the commonly used personalized learning methods. A web crawler is used to crawl English exercises data on the Internet and classify them according to the knowledge categories, establish the connection between topics according to their cosine similarity, and build an English knowledge graph using NEO4J. After creating the knowledge map, a select few nodes are categorized with labels that reflect the characteristics of various pupils, and the node categories are divided into two categories: mastered and unmastered, after which the whole knowledge map is labeled using a graph convolutional neural network to generate a personalized knowledge map. After testing, the model works best when the threshold value is increased to 0.8 when generating the English knowledge map. And the overall English score of the experimental class taught with this paper’s method improved by 5.27 points, among which the average scores of completion, writing, and reading increased by 2.58, 1.32, and 1.17 points, while the overall English level of the control class did not change significantly. As a result, the approach used in this research to create tailored learning routes based on personalized knowledge mapping has excellent applicability.</p>
</abstract>

This paper proposes a personalized teaching method based on knowledge mapping and the commonly used personalized learning methods. A web crawler is used to crawl English exercises data on the Internet and classify them according to the knowledge categories, establish the connection between topics according to their cosine similarity, and build an English knowledge graph using NEO4J. After creating the knowledge map, a select few nodes are categorized with labels that reflect the characteristics of various pupils, and the node categories are divided into two categories: mastered and unmastered, after which the whole knowledge map is labeled using a graph convolutional neural network to generate a personalized knowledge map. After testing, the model works best when the threshold value is increased to 0.8 when generating the English knowledge map. And the overall English score of the experimental class taught with this paper’s method improved by 5.27 points, among which the average scores of completion, writing, and reading increased by 2.58, 1.32, and 1.17 points, while the overall English level of the control class did not change significantly. As a result, the approach used in this research to create tailored learning routes based on personalized knowledge mapping has excellent applicability.

A Study on Innovative Strategies for Teaching English to College Students Based on Knowledge Mapping

Applied Mathematics and Nonlinear Sciences

This work is licensed under the Creative Commons Attribution 4.0 International License.

This paper proposes a personalized teaching method based on knowledge mapping and the commonly used personalized learning methods. A web crawler is used to...

A Study on Innovative Strategies for Teaching English to College Students Based on Knowledge Mapping

Publicado en línea: 16 sept 2023

Páginas: -

Recibido: 18 dic 2022

Aceptado: 27 mar 2023

DOI: https://doi.org/10.2478/amns.2023.2.00339

Palabras clave
NEO4J, Knowledge graph, Graph convolutional neural network, Personalized learning, Cosine similarity

© 2023 Lun Wang, published by Sciendo

This work is licensed under the Creative Commons Attribution 4.0 International License.

A Study on Innovative Strategies for Teaching English to College Students Based on Knowledge Mapping

Publicado en línea: 16 sept 2023

Páginas: -

Recibido: 18 dic 2022

Aceptado: 27 mar 2023

DOI: https://doi.org/10.2478/amns.2023.2.00339

Palabras claveNEO4J, Knowledge graph, Graph convolutional neural network, Personalized learning, Cosine similarity

© 2023 Lun Wang, published by Sciendo

This work is licensed under the Creative Commons Attribution 4.0 International License.

Palabras clave
NEO4J, Knowledge graph, Graph convolutional neural network, Personalized learning, Cosine similarity