A Pluggable Multi-Task Learning Framework for Sentiment-Aware Financial Relation Extraction

2025-06-14

AI Summary

Sentiment information has been long overlooked in financial relation extraction, leading to insufficient semantic modeling. Method: This paper proposes SSDP-SEM, a plug-and-play multi-task framework that jointly models relation extraction and fine-grained sentiment perception. It introduces the first collaborative modeling mechanism integrating sentiment signals with shortest dependency paths (SDPs) and designs a sentiment-aware attention information bottleneck regularization for controllable, sentiment-guided inference. SSDP-SEM is compatible with mainstream RE models (e.g., BERT, SpanBERT) via sentiment token injection, dependency-guided attention, and multi-task optimization. Contribution/Results: Experiments demonstrate consistent and significant F1-score improvements (+2.1–3.8 points) across multiple financial relation datasets. This work provides the first systematic empirical validation of the critical role sentiment signals play in enhancing domain-specific relation modeling.

Technology Category

• Natural Language Processing: Information Extraction
• Reasoning under Uncertainty: Relational Probabilistic Models
• Machine Learning: Statistical Relational/Logic Learning

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Web Mining and Content Analysis: Sentiment analysis and opinion mining
• Graph Algorithms and Modeling for the Web: Algorithms and analysis for heterogeneous, signed, attributed, multi-relational, temporal, higher-order, and annotated Web-related graphs

Abstract

Relation Extraction (RE) aims to extract semantic relationships in texts from given entity pairs, and has achieved significant improvements. However, in different domains, the RE task can be influenced by various factors. For example, in the financial domain, sentiment can affect RE results, yet this factor has been overlooked by modern RE models. To address this gap, this paper proposes a Sentiment-aware-SDP-Enhanced-Module (SSDP-SEM), a multi-task learning approach for enhancing financial RE. Specifically, SSDP-SEM integrates the RE models with a pluggable auxiliary sentiment perception (ASP) task, enabling the RE models to concurrently navigate their attention weights with the text's sentiment. We first generate detailed sentiment tokens through a sentiment model and insert these tokens into an instance. Then, the ASP task focuses on capturing nuanced sentiment information through predicting the sentiment token positions, combining both sentiment insights and the Shortest Dependency Path (SDP) of syntactic information. Moreover, this work employs a sentiment attention information bottleneck regularization method to regulate the reasoning process. Our experiment integrates this auxiliary task with several prevalent frameworks, and the results demonstrate that most previous models benefit from the auxiliary task, thereby achieving better results. These findings highlight the importance of effectively leveraging sentiment in the financial RE task.

Problem

• Enhancing financial relation extraction with sentiment awareness
• Integrating sentiment perception into relation extraction models
• Improving RE accuracy using sentiment and syntactic information

Innovation

• Pluggable multi-task learning framework
• Sentiment-aware-SDP-Enhanced-Module (SSDP-SEM)
• Sentiment attention information bottleneck regularization