Implicit discourse relation recognition (IDRR) still faces challenges in achieving high accuracy compared to explicit discourse relation recognition. Many existing approaches attempt to combine explicit and implicit data. However, language differences and variations in category distribution make it challenging to use multitask learning to directly optimize IDRR task. This paper considers explicit data as out-of-domain data and implicit data as in-domain data, proposed a shared-private-based framework with dynamic dependency matching (SPDDM) for IDRR task. This framework explores the deep-level syntactic correlation between explicit and implicit data by encoding dependency grammar information through cross-attention mechanism and confuse it to shared-private framework through dynamic matching . Extensive experiments on the popular PDTB datasets show that our model achieves state-of-the-art performance, and a detailed analysis highlights the efficiency of our proposed framework.

This paper proposes incorporating linguistic semantic information into discourse relation recognition and constructing a Semantic Augmented Chinese Discourse Corpus (SACA) comprising 9546 adversative complex sentences. In adversative complex sentences, we suggest a quadruple (P, Q, R, $Q_{\beta }$) representing internal semantic elements, where the semantic opposition between Q and $Q_{\beta }$ forms the basis of the adversative relationship. P denotes the premise, and R represents the adversative reason. The overall annotation approach of this corpus follows the Penn Discourse Treebank (PDTB), except for the classification of senses. We combined insights from the Chinese Discourse Treebank (CDTB) and obtained eight sense categories for Chinese adversative complex sentences. Based on this corpus, we explore the relationship between sense classification and internal semantic elements within our newly proposed Chinese Adversative Discourse Relation Recognition (CADRR) task. Leveraging deep learning techniques, we constructed various classification models and the model that utilizes internal semantic element features, demonstrating their effectiveness and the applicability of our SACA corpus. Compared with pre-trained models, our model incorporates internal semantic element information to achieve state-of-the-art performance.