Fine-Grained Sentiment-Controlled Text Generation Approach Based on Pre-Trained Language Model

Article Computer Science, Artificial Intelligence

CenGCN: Centralized Convolutional Networks with Vertex Imbalance for Scale-Free Graphs

Feng Xia et al.

Summary: Graph Convolutional Networks (GCNs) have achieved impressive performance in various areas and have attracted considerable attention. However, the equal importance given to all information in the information-passing framework of GCNs is insufficient for scale-free networks. This paper proposes a centrality-based framework called CenGCN to address this inequality. By quantifying the similarity between hub vertices and their neighbors, the framework transforms the graph and assigns new weights to non-hub vertices based on their common information with hub vertices. Experimental results show that the proposed framework outperforms state-of-the-art baselines in tasks such as vertex classification, link prediction, vertex clustering, and network visualization.

IEEE TRANSACTIONS ON KNOWLEDGE AND DATA ENGINEERING (2023)

添加到收藏夹

Review Computer Science, Artificial Intelligence

Deep learning for aspect-based sentiment analysis: a review

Linan Zhu et al.

Summary: This article provides an overview of deep learning for aspect-based sentiment analysis. It introduces the ABSA task and presents the overall framework from the perspectives of significant subtasks and the task modeling process. The challenges in sentiment analysis, particularly in the field of aspect-based sentiment analysis, are discussed. Additionally, the ABSA task's consideration of relations between various objects is highlighted.

PEERJ COMPUTER SCIENCE (2022)

添加到收藏夹

Article Computer Science, Information Systems

Hierarchical template transformer for fine-grained sentiment controllable generation

Li Yuan et al.

Summary: This paper proposes a hierarchical template-transformer model for personalized fine-grained sentiment controllable generation. The model utilizes a hierarchical structure and part of speech (POS) templates to generate aspect-level sentiment controllable reviews with personalized information.

INFORMATION PROCESSING & MANAGEMENT (2022)

添加到收藏夹

Proceedings Paper Computer Science, Information Systems

Mutual Disentanglement Learning for Joint Fine-Grained Sentiment Classification and Controllable Text Generation

Hao Fei et al.

Summary: This paper combines fine-grained sentiment classification and fine-grained controllable text generation as a joint dual learning system, improving the performance of both tasks through mutual learning. By decoupling the feature representations and transforming the difficult data-to-text generation into an easier text-to-text generation, the model's effectiveness is further enhanced.

PROCEEDINGS OF THE 45TH INTERNATIONAL ACM SIGIR CONFERENCE ON RESEARCH AND DEVELOPMENT IN INFORMATION RETRIEVAL (SIGIR '22) (2022)

添加到收藏夹

Proceedings Paper Computer Science, Artificial Intelligence

Controllable Generation from Pre-trained Language Models via Inverse Prompting

Xu Zou et al.

Summary: Inverse prompting is an innovative method proposed to better control text generation, which uses generated text to inversely predict the prompt during beam search, enhancing the relevance between the prompt and the generated text to improve controllability. Empirical results show that this method outperforms baselines significantly and achieves generation quality close to human performance on certain tasks.

KDD '21: PROCEEDINGS OF THE 27TH ACM SIGKDD CONFERENCE ON KNOWLEDGE DISCOVERY & DATA MINING (2021)

添加到收藏夹

Fine-Grained Sentiment-Controlled Text Generation Approach Based on Pre-Trained Language Model

相关参考文献

CenGCN: Centralized Convolutional Networks with Vertex Imbalance for Scale-Free Graphs

Deep learning for aspect-based sentiment analysis: a review

Hierarchical template transformer for fine-grained sentiment controllable generation

Mutual Disentanglement Learning for Joint Fine-Grained Sentiment Classification and Controllable Text Generation

Controllable Generation from Pre-trained Language Models via Inverse Prompting

导出引文

分享论文