AntCritic: Argument Mining for Free-Form and Visually-Rich Financial Comments

TL;DR

This paper introduces AntCritic, a large, multimodal dataset of visually-rich financial comments for argument mining, along with models and benchmarks to advance research in argument component detection and relation prediction.

Contribution

The paper presents a novel, large-scale dataset with visual and textual data for argument mining, and proposes effective models and benchmarks for this multimodal task.

Findings

AntCritic dataset contains about 10,000 comments with visual and textual information.

Proposed models achieve competitive benchmark performance on argument component detection.

The dataset and models facilitate future research in multimodal argument mining.

Abstract

Argument mining aims to detect all possible argumentative components and identify their relationships automatically. As a thriving task in natural language processing, there has been a large amount of corpus for academic study and application development in this field. However, the research in this area is still constrained by the inherent limitations of existing datasets. Specifically, all the publicly available datasets are relatively small in scale, and few of them provide information from other modalities to facilitate the learning process. Moreover, the statements and expressions in these corpora are usually in a compact form, which restricts the generalization ability of models. To this end, we collect a novel dataset AntCritic to serve as a helpful complement to this area, which consists of about 10k free-form and visually-rich financial comments and supports both argument component detection and argument relation prediction tasks. Besides, to cope with the challenges brought by scenario expansion, we thoroughly explore the fine-grained relation prediction and structure reconstruction scheme and discuss the encoding mechanism for visual styles and layouts. On this basis, we design two simple but effective model architectures and conduct various experiments on this dataset to provide benchmark performances as a reference and verify the practicability of our proposed architecture. We release our data and code in this link, and this dataset follows CC BY-NC-ND 4.0 license.