Knowledge-guided Transformer for Joint Theme and Emotion Classification of Chinese Classical Poetry

Abstract

The classifications of the theme and emotion are essential for understanding and organizing Chinese classical poetry. Existing works fail to consider the lexical knowledge mined from poem annotations, which intuitively reflects the theme and emotion. In addition, they just treat them as two separate tasks without considering that the emotion is usually related with the theme. In this paper, we propose a Knowledge-guided Transformer Model (KTM) for joint theme and emotion classification of Chinese classical poetry. Specifically, we first respectively construct two lexical dictionaries for the theme and emotion based on the poem annotations. Then we take full advantage of the lexical dictionaries with a knowledge-based mask-transformer to represent poems. Furthermore, considering the correlations between the theme and emotion, our model jointly classifies the theme and emotion for Chinese classical poetry by stacking the two subtasks. Extensive experiments demonstrate that our model achieves state-of-the-art performance on both theme and emotion classifications, especially on tail labels.

Tasks

Reproductions