Hierarchical Macro Discourse Parsing Based on Topic Segmentation

Abstract

Hierarchically constructing micro (i.e., intra-sentence or inter-sentence) discourse structure trees using explicit boundaries (e.g., sentence and paragraph boundaries) has been proved to be an effective strategy. However, it is difficult to apply this strategy to document-level macro (i.e., inter-paragraph) discourse parsing, the more challenging task, due to the lack of explicit boundaries at the higher level. To alleviate this issue, we introduce a topic segmentation mechanism to detect implicit topic boundaries and then help the document-level macro discourse parser to construct better discourse trees hierarchically. In particular, our parser first splits a document into several sections using the topic boundaries that the topic segmentation detects. Then it builds a smaller and more accurate discourse sub-tree in each section and sequentially forms a whole tree for a document. The experimental results on both Chinese MCDTB and English RST-DT show that our proposed method outperforms the state-of-the-art baselines significantly.

It’s Nice to Meet You: Contextual Configuration on Formal and Informal Introductions

“…The theory of contextual configuration emerges as an approach to analyzing the structural integrity of a text or discourse. Discourse necessarily has an element of impartial coherence in its structure (Dijk, 2019;Jiang et al, 2021). Certainly, the structure in this regard leads to the sequence or syntagmatic relation of discourse units or parts-e.g., in the Aristotelian tradition, the discourse structure is divided into three major parts, viz.…”

Opinion mining on product reviews from document to clause

“…Elementary discourse units are clause-like grammatical units that serve as building blocks for discourse parsing in rhetorical structure theory (RST) [24]. In studies on topic segmentation based on RST, EDU is considered as a basic semantic unit; it carries coherent semantic meaning towards a subtopic in text [132,133,134,135,136,137,138,139]. Applying the same intuition to a complex sentence, we observe that an EDU bears unitary polarity toward a single aspect, as shown in Table 4.1.…”