CRF-GCN: An effective syntactic dependency model for aspect-level sentiment analysis

Article Computer Science, Artificial Intelligence

Meta-Based Self-Training and Re-Weighting for Aspect-Based Sentiment Analysis

Kai He et al.

Summary: The study proposes a meta-based self-training method for aspect-based sentiment analysis (ABSA). By generating pseudo-labels and controlling convergence rates, the method improves model performance and accuracy in fine-grained sentiment analysis.

IEEE TRANSACTIONS ON AFFECTIVE COMPUTING (2023)

Add to Collection

Article Computer Science, Artificial Intelligence

Aspect-based sentiment analysis via affective knowledge enhanced graph convolutional networks

Bin Liang et al.

Summary: This paper proposes a graph convolutional network model Sentic GCN based on SenticNet to enhance the affective dependencies of sentences for aspect-based sentiment analysis. By integrating emotional knowledge from SenticNet, the model effectively handles contextual affective information in sentences, improving the effectiveness of sentiment polarity detection towards specific aspects.

KNOWLEDGE-BASED SYSTEMS (2022)

Add to Collection

Article Computer Science, Artificial Intelligence

Aspect-level sentiment analysis with aspect-specific context position information

Bo Huang et al.

Summary: In recent years, researchers have paid more attention to aspect-level sentiment analysis in natural language processing. A fine-grained sentiment analysis distinguishes each aspect of the text and makes separate judgments on the sentiment polarity. This paper proposes an aspect-level sentiment analysis model with aspect-specific contextual location information, adjusting the weight of contextual words and extracting the influence of contextual association on individual sentence sentiment polarity.

KNOWLEDGE-BASED SYSTEMS (2022)

Add to Collection

Proceedings Paper Computer Science, Interdisciplinary Applications

Neuro-Symbolic Models for Sentiment Analysis

Jan Kocon et al.

Summary: In this study, we propose and test multiple neuro-symbolic methods for sentiment analysis. We find that adding information from knowledge bases significantly improves the quality of sentiment prediction for simple models. Additionally, our proposed methods, Tailored KEPLER and Token Extension, show significant improvements over state-of-the-art transformer-based models for medium-sized datasets.

COMPUTATIONAL SCIENCE, ICCS 2022, PT II (2022)

Add to Collection

Article Computer Science, Artificial Intelligence