基于表示学习的跨学科概念关联研究OA

[Purpose/Significance]To address the challenges in deep representation learning of concepts within multi-level semantic relationships and their interdisciplinary associations,and to break through the limitations of traditional me-thods that rely solely on surface feature matching.[Method/Process]This paper proposes an interdisciplinary concept associa-tion method based on a subject concept knowledge graph.This method leverages a representation learning-based know-ledge alignment model to integrate syntactic,semantic,and pragmatic correlations,capturing the implicit structural asso-ciation features within the subject knowledge graph to construct a concept association model oriented towards interdiscipli-nary knowledge services.[Result/Conclusion]This paper tests the proposed method using the field of"privacy protection"as the experimental subject,and validates the effectiveness of the interdisciplinary concept association method based on representation learning.