Abstract:Aspect-based sentiment analysis is a fine-grained task where the key goal is to predict sentiment polarities of one or more aspects in a given sentence. Currently, graph neural network models built upon dependency trees are widely employed for aspect-based sentiment analysis tasks. However, most existing models still contain a large amount of noisy nodes that cannot precisely capture the contextual relationships between specific aspects. Meanwhile, most studies do not consider the connections between nodes without direct dependency edges but play critical roles in determining the sentiment polarity of an aspect. To address the aforementioned limitations, we propose a Structured Dependency Tree-based Graph Convolutional Network (SDTGCN) model. Specifically, we explore construction of a structured syntactic dependency graph by incorporating positional information, sentiment commonsense knowledge, part-of-speech tags, syntactic dependency distances, etc., to assign arbitrary edge weights between nodes. This enhances the connections between aspect nodes and pivotal words while weakening irrelevant node links, enabling the model to sufficiently express sentiment dependencies between specific aspects and contextual information. We utilize part-of-speech tags and dependency distances to discover relationships between pivotal nodes without direct dependencies. Finally, we aggregate node information by fully considering their importance to obtain precise aspect representations. Experimental results on five publicly available datasets demonstrate the superiority of our proposed model over state-of-the-art approaches; furthermore, the accuracy and F1-score show a significant improvement on the majority of datasets, with increases of 0.74, 0.37, 0.65, and 0.79, 0.75, 1.17, respectively. This series of enhancements highlights the effective progress made by the STDGCN model in enhancing sentiment classification performance.

What problem does this paper attempt to address?

### The Problem Addressed by This Paper This paper primarily focuses on the fine-grained sentiment analysis task—Aspect-Based Sentiment Analysis (ABSA), aiming to predict the sentiment polarity (such as positive, neutral, negative) of specific aspects within a sentence. Currently, graph neural network models constructed based on dependency trees are widely used in ABSA tasks, but they have the following issues: 1. **Noise Nodes**: Existing models contain a large number of noise nodes, failing to accurately capture contextual relationships. 2. **Non-Direct Dependency Relationships**: Most studies ignore nodes that do not have direct dependency relationships but are crucial for sentiment polarity judgment. 3. **Dependency Tree Weights**: The weights of edges in existing dependency trees are binary, failing to distinguish the strength of relationships between nodes. To address these issues, the paper proposes a new model—Structured Dependency Tree-based Graph Convolutional Network (SDTGCN). Specifically, this model reconstructs the edge weights of the dependency tree by incorporating positional information, sentiment commonsense knowledge, part-of-speech tags, and syntactic dependency distance. This enhances the connection between specific aspects and key terms while weakening the connections of irrelevant nodes, thereby better expressing the sentiment dependency relationships and contextual information of specific aspects. Experimental results show that this model outperforms existing methods on 5 public datasets, with significant improvements in both accuracy and F1 score.

Modeling Structured Dependency Tree with Graph Convolutional Networks for Aspect-Level Sentiment Classification

Dual Edge-embedding Graph Convolutional Network for Unified Aspect-based Sentiment Analysis

Graph convolutional network with interactive memory fusion for aspect-based sentiment analysis

Modeling sentiment dependencies with graph convolutional networks for aspect-level sentiment classification

Aspect-based Sentiment Classification with Aspect-specific Graph Convolutional Networks

A Relative Position Attention Network for Aspect-Based Sentiment Analysis

Syntactic and Semantic Aware Graph Convolutional Network for Aspect-Based Sentiment Analysis

Graph Convolution over the Semantic-syntactic Hybrid Graph Enhanced by Affective Knowledge for Aspect-level Sentiment Classification

Convolution over Hierarchical Syntactic and Lexical Graphs for Aspect Level Sentiment Analysis

Graph Convolutional Network with Syntactic Dependency for Aspect-Based Sentiment Analysis

Hybrid Graph Neural Network-Based Aspect-Level Sentiment Classification

RDGCN: Reinforced Dependency Graph Convolutional Network for Aspect-based Sentiment Analysis

Dual-Channel Interactive Graph Convolutional Networks for Aspect-Level Sentiment Analysis

Aggregated graph convolutional networks for aspect-based sentiment classification

Exploring fine-grained syntactic information for aspect-based sentiment classification with dual graph neural networks

A Hybrid Semantics and Syntax-Based Graph Convolutional Network for Aspect-Level Sentiment Classification

Syntactic Structure-Enhanced Dual Graph Convolutional Network for Aspect-Level Sentiment Classification

Aspect-level sentiment classification with aspect-opinion sentence pattern connection graph convolutional networks

Twain-GCN: twain-syntax graph convolutional networks for aspect-based sentiment analysis

Selective Attention Based Graph Convolutional Networks for Aspect-Level Sentiment Classification

Multiple graph convolutional networks for aspect-based sentiment analysis