TL;DR
The Syntactic-optimal Transport Graph Network (SOT-Graph) is proposed, a model that jointly integrates structural and distributional signals and outperforms existing baselines by a margin of 1.30% Macro-F1 on Laptop14 and 1.01% on Twitter.
Recommendation systems are rapidly evolving from static interaction‐driven models to dynamic, knowledge‐augmented architectures. A key challenge in this evolution is accurately capturing users' fine‐grained preferences from unstructured review text, which directly impacts the explainability and personalization of recommendations. As an essential enabling technology, Aspect‐Based Sentiment Analysis (ABSA) extracts aspect‐level sentiment elements that can be explicitly mapped to user preference vectors or product attribute ratings. With the integration of semantic and syntactic information, current works have significantly enhanced the performance of ABSA. However, existing graph‐based approaches that rely on dependency‐tree structures often converge to suboptimal solutions when handling implicit sentiment in natural language. To address this gap, we propose a graph fusion network that leverages augmented semantic and syntactic graphs. Specifically, we explicitly model word‐dependency correlations via contextual augmentation, and incorporate selected part‐of‐speech (POS) features to refine semantic graph construction. Concurrently, a syntactic graph is constructed by pruning the nodes based on the distance to the aspect term. The resulting semantic and syntactic representations are then fused through a dual graph convolutional network block, whereas the gating mechanism is used to regulate information flow during graph construction. Experiments on seven benchmarks demonstrate that our approach outperforms baselines by up to 2.39%$$ 2.39\% $$ and 2.44%$$ 2.44\% $$ in Macro‐F1 scores, establishing new state‐of‐the‐art results and providing a more reliable sentiment extraction module for downstream recommendation tasks.
Aspect-Based Sentiment Analysis (ABSA) aims to determine the sentiment polarity expressed toward specific aspect terms within a sentence. Existing approaches often rely on syntactic dependency parse graphs to model the relationship between aspect terms and opinion words. Despite recent progress, two key challenges remain unresolved. First, dependency parse graphs often contain task-irrelevant yet syntactically valid edges that can mislead sentiment reasoning. Second, meaningful sentiment relations between aspect terms and opinion words frequently span multiple dependency edges, making it difficult for models to capture long-range semantic interactions. To address these challenges, we propose a Sentiment-Aware Graph Anomaly detection and Mix convolution (SAGAM) model. The sentiment-aware graph anomaly detection framework integrates graph anomaly detection with a sentiment consistency evaluation mechanism to eliminate misleading dependencies and preserve key sentiment information from the original graph. To further capture long-range dependencies between aspect terms and opinion words, the mix convolution module combines multi-hop message propagation with edge feature integration, enabling the model to capture both local and distant semantic interactions. Extensive experiments on three widely used ABSA benchmark datasets demonstrate that SAGAM consistently outperforms strong baselines in both accuracy and macro-F1. Notably, the proposed sentiment-aware graph anomaly detection framework can be seamlessly incorporated into other graph-based ABSA models to enhance performance and robustness.
Aspect-based sentiment analysis (ABSA) demands more than a single polarity label per document — it requires assigning a distinct sentiment to each named attribute a text discusses, even when the relevant opinion words are syntactically distant, negated, or figurative. Existing graph convolutional approaches to ABSA rely on three design decisions that limit how far they can go: sentiment lexicon weights computed once before training that cannot adapt to how words are used in context; a single deterministic dependency parser whose errors embed directly into the adjacency matrix with no correction path; and graphs built exclusively from token-level nodes, leaving clause-level and phrase-level sentiment units without any dedicated representation. MH-ABSA addresses each of these gaps through five independently ablatable modules. LLM-Guided Dynamic Edge Reweighting (LDEM) substitutes static lexicon scores with context-sensitive weights produced by a frozen instruction-tuned language model, blended into the adjacency matrix through a learnable coefficient. Ensemble-Uncertainty Dependency Fusion (EUDF) combines three structurally diverse parsers through a sentence-conditioned MLP, allowing the model’s trust in each parser to adapt with the input register. The Multi-Scale Heterogeneous Graph (MHG) extends the token-level graph to three tiers — token, span, and discourse — so that clause-level sentiment signals can propagate to individual aspect tokens in a single message-passing hop rather than through a long token chain. Two further modules round out the design. Contrastive Sentiment Alignment Module (CSAM) introduces a contrastive training objective built around LLM-generated paraphrases and domain-internal aspect swaps, giving the model a second gradient pathway that operates on meaning rather than surface vocabulary — useful precisely because sarcasm and implicit sentiment defeat both the lexicon and the parser. Domain-Adaptive Edge Masking (DAEM) addresses a subtler problem: polarity scores that are globally reasonable but locally wrong for a given product category, which it handles through per-edge learned gates that suppress unreliable weights at training time. The proposed framework addresses several limitations of existing graph-based ABSA systems, including static lexicon weighting, parser uncertainty, and lack of multi-scale linguistic representations.
Aspect-Based Sentiment Analysis (ABSA) aims to identify opinion targets and determine the sentiment polarity expressed toward each aspect, which requires fine-grained modeling of aspect–opinion relations. Despite recent advances, cross-domain ABSA remains challenging due to structural mismatches across domains and the scarcity of high-quality labeled data in target domains. Existing methods often struggle to jointly address relational modeling errors and data sparsity, particularly under low-resource and cross-domain settings. To tackle these challenges, we propose a structure- and data-co-enhanced framework for cross-domain ABSA. At the model level, we introduce an Enhanced Affinity Fusion (EAF) module that explicitly strengthens aspect–opinion relational modeling by selectively integrating complementary attention mechanisms. Specifically, EAF combines biaffine attention to capture second-order interactions with syntax-aware attention to inject structural inductive bias, enabling robust modeling of long-distance dependencies without introducing excessive architectural complexity. At the data level, we propose Label-Guided Data Amplification (LGDA), which enhances supervision diversity and domain robustness through label-driven text expansion, hard sample mining, and domain-adaptive sampling. By jointly enhancing structural representation learning and training data supervision, the proposed framework effectively alleviates both aspect–opinion mismatches and cross-domain data sparsity. Extensive experiments on benchmark ABSA datasets demonstrate that our approach consistently outperforms strong baselines and achieves state-of-the-art performance in cross-domain scenarios. Ablation studies further validate the complementary contributions of EAF and LGDA.
Aspect-Based Sentiment Analysis (ABSA) often experiences a significant performance decline in cross-domain settings due to vocabulary variation and domain-specific aspect expressions. Although transformer-based models achieve strong in-domain performance, they primarily rely on contextual embeddings and often ignore the syntactic structures that remain consistent across domains. Existing methods rarely integrate structured decoding with adaptive syntactic fusion for robust aspect boundary detection. This paper proposes a syntactic-aware cross-domain ABSA framework based on DeBERTaV3 and BIO-CRF decoding to alleviate the above problems. The proposed model introduces part-of-speech and dependency-relation embeddings, in addition to contextual embeddings, and uses an attention-based model to dynamically fuse syntactic and semantic information at multiple levels. We further apply a Conditional Random Field (CRF) layer to enforce valid BIO transitions and enhance the consistency of multi-word aspect spans under domain shift. The model was evaluated in three English review domains: Restaurant, Laptop, and Device across six zero-shot cross-domain transfer settings (D→L, D→R, L→D, L→R, R→D, and R→L). Test results show consistent advances over robust transformer-based and prompt-based baselines. The proposed method yields F1 scores for aspect extraction between 0.72 and 0.81 and achieves sentiment classification accuracies between 74.32% and 85.19%. The best performance was achieved in the L→R transfer setting. Through paired bootstrap testing (p < 0.01), Statistical analysis confirms that the proposed model achieves significant improvements over baseline methods. The results demonstrate that incorporating explicit syntactic knowledge, adaptive feature fusion, and structured decoding substantially improves cross-domain generalization in fine-grained sentiment analysis.
Aspect-based Sentiment Analysis (ABSA) is a fine-grained sentiment classification task that aims to predict the sentiment polarity associated with aspect terms in sentences. Traditional methods based on syntactic and semantic dependency trees are insufficient for capturing contextual sentence features. To address this, we propose a Knowledge-Enhanced and Multidimensional Feature Fusion (KEMFF) model for ABSA, which captures sentiment feature representations across multiple dimensions, including syntax, semantics, and knowledge. First, the pre-trained model RoBERTa is used to obtain embeddings of sentences and aspect terms. Then, a syntactic dependency parser and a graph convolutional network are utilized to learn syntactic features. Meanwhile, an Abstract Meaning Representation (AMR)-based parser is employed to construct semantic relations, and axial attention is used to aggregate incoming and outgoing semantic dependencies. Furthermore, external knowledge is embedded, and an attention mechanism is employed to obtain aspect-specific knowledge representations, thereby complementing syntactic and semantic representations with external lexical knowledge. Finally, multidimensional features are fused and passed to a softmax classifier for predicting sentiment polarities. Unlike previous models that mainly focus on either syntax–semantic fusion or knowledge-enhanced graph propagation, KEMFF explicitly models syntax, semantics, and lexical knowledge in three parallel branches and aligns them into a unified aspect-level representation. Experiments on Laptop14, Restaurant14, and Twitter datasets show that KEMFF achieves the best performance among the compared baselines on Laptop14 and Restaurant14, and it obtains competitive results on Twitter.