Learning Term Embeddings for Taxonomic Relation Identification with Dynamic Weighting Neural Network

Abstract

In this paper, the aim is to recognize the ‘is-a’ relation between two terms through taxonomic relation identification. Previous efforts to identify taxonomic relations have primarily relied on statistical and linguistic approaches, but the accuracy of these methods is far from satisfactory. To address this, we propose a novel supervised learning approach that utilizes term embeddings. To achieve this, we first design a dynamic weighting neural network to learn term embeddings, considering not only the hypernym and hyponym terms but also the contextual information between them. Subsequently, we apply these embeddings as features in a supervised method for identifying taxonomic relations. The experimental results demonstrate that our proposed approach significantly outperforms other state-of-the-art methods, achieving a 9% to 13% increase in accuracy for both general and specific domain datasets.

Luu Anh Tuan

Assistant Professor

My research interests lie in the intersection of Artificial Intelligence and Natural Language Processing.