.png){kind=icon}
.png){kind=icon}
Return
Multi-Task Learning With LLMs for Implicit Sentiment Analysis: Data-Level and Task-Level Automatic Weight Learning
.png){kind=icon}
OA
AI
DOI：10.1109/TKDE.2025.3623941.png)
.png){kind=icon}
.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Implicit sentiment analysis (ISA) presents significant challenges due to the absence of salient cue words. Previous methods have struggled with insufficient data and limited reasoning capabilities to infer underlying opinions. Integrating multi-task learning (MTL) with large language models (LLMs) offers the potential to enable models of varying sizes to reliably perceive and recognize genuine opinions in ISA. However, existing MTL approaches are constrained by two sources of uncertainty: <bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><i>data-level uncertainty</i></b>, arising from hallucination problems in LLM-generated contextual information, and <bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><i>task-level uncertainty</i></b>, stemming from the varying capacities of models to process contextual information. To handle these uncertainties, we propose <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">MT-ISA</i>, a novel MTL framework that enhances ISA by leveraging the generation and reasoning capabilities of LLMs through automatic weight learning (AWL). Specifically, <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">MT-ISA</i> constructs auxiliary tasks using generative LLMs to supplement sentiment elements and incorporates automatic MTL to fully exploit auxiliary data. We introduce data-level and task-level AWL, which dynamically identify relationships and prioritize more reliable data and critical tasks, enabling models of varying sizes to adaptively learn fine-grained weights based on their reasoning capabilities. Three strategies are investigated for data-level AWL, which are integrated with homoscedastic uncertainty for task-level AWL. Extensive experiments reveal that models of varying sizes achieve an optimal balance between primary prediction and auxiliary tasks in <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">MT-ISA</i>. This underscores the effectiveness and adaptability of our approach.
Keywords:
Implicit sentiment analysis
multi-task learning
large language models
AI Summary
Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.
Journal
IF:
10.4
Papers:
6.7K
Citations:
3.2W
Organization
View more organizations