Multimodal Sentiment Analysis Based on Cross-Modal Attention and Gated Cyclic Hierarchical Fusion Networks

Quan, Zhibang; Sun, Tao; Su, Mengli; Wei, Jishu

doi:10.1155/2022/4767437

Cited by 8 publications

(5 citation statements)

References 30 publications

(40 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In this study, STEP length is set to 1, and three convolution kerns of different sizes (3,4,5) are used to extract local features of sentences 3-gram, 4-gram and 5-gram respectively. To accelerate training convergence, the ReLU function serves as the activation function in this study, with its mathematical expression shown in formula (3).…”

Section: Figure 3 Bert Model Structurementioning

confidence: 99%

“…Depending on the number of emotional categories involved, the emotion tendency analysis tasks are categorized into two groups(positive, negative), three groups (positive, negative, neutral) and multiple groups (happy, excited, sad, angry, etc.) [3]. Analysis of textual emotion tendency mainly contains: text representation and feature extraction, model training and analysis of the results [4].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Text Sentiment Analysis of Douban Film Short Comments Based on BERT-CNN-BiLSTM-Att Model

He,

Abisado

2024

IEEE Access

View full text Add to dashboard Cite

To solve the problems of polysemy and feature extraction in the text sentiment analysis process, a BERT-CNN-BiLSTM-Att hybrid model has been proposed for text sentiment analysis. The BERT pretraining model was established to break up the text input into words and obtain a dynamic word vector that was then input into the CNN and the BiLSTM models respectively. Later, the local features of the word vector, extracted using CNN, and the global features, extracted using BiLSTM, were fused, and the key information of the Douban movie review dataset was highlighted using the attention mechanism to realize sentiment categorization of the dataset. The results of comparison between the constructed model and Word2Vec-BiLSTM,Word2Vec-CNN,Word2Vec-CNN-BiLSTM-Att,BERT,BERT-CNN and BERT-BiLSTM models show that the model that runs against the test dataset has an increased accuracy by 4.63%,4.37%,3.64%,2.63%,2.56% and 5.54% respectively. The experimental findings reveal that BERT-CNN-BiLSTM-Att's sentiment analysis method is more accurate in performing sentiment classification.INDEX TERMS BERT-CNN-BiLSTM-Att, Sentiment analysis, Hybrid model, Film short text reviews Comments

show abstract

Section: Figure 3 Bert Model Structurementioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Text Sentiment Analysis of Douban Film Short Comments Based on BERT-CNN-BiLSTM-Att Model

He,

Abisado

2024

IEEE Access

View full text Add to dashboard Cite

show abstract

“…Wearable robots have existed for many years, but it is only in recent years that significant breakthroughs have occurred. Multi-modal fusion 13 , human-in-the-loop control 14 , neuromuscular interface 15 , flexible electronics 16 , and biomechatronic chip 17 , represent examples of advancements that could profoundly and positively impact the interaction between wearable robots and humans (Table 1 ). However, due to their complex nature, only a few of these advancements have been tested with users, often involving a limited number of participants 18 .…”

Section: Introductionmentioning

confidence: 99%

“…Biomechatronic chips 127 could serve as the central unit for information acquisition and processing, generating control commands. Neuromuscular interface 128 and flexible electronics 129 are enabling technologies for sensing human intention, which could be fused together with multi-modal fusion 13 . They also provide sensory feedback that transfers information from the robot to the human to improve the feeling of agency.…”

Section: Introductionmentioning

confidence: 99%

Shaping high-performance wearable robots for human motor and sensory reconstruction and enhancement

Xia,

Zhang,

Rajabi

et al. 2024

Nat Commun

View full text Add to dashboard Cite

Most wearable robots such as exoskeletons and prostheses can operate with dexterity, while wearers do not perceive them as part of their bodies. In this perspective, we contend that integrating environmental, physiological, and physical information through multi-modal fusion, incorporating human-in-the-loop control, utilizing neuromuscular interface, employing flexible electronics, and acquiring and processing human-robot information with biomechatronic chips, should all be leveraged towards building the next generation of wearable robots. These technologies could improve the embodiment of wearable robots. With optimizations in mechanical structure and clinical training, the next generation of wearable robots should better facilitate human motor and sensory reconstruction and enhancement.

show abstract

“…Many large knowledge graphs, such as YAGO [ 2 ], Freebase [ 3 ], and DBpedia [ 4 ], use triples to store the entities and relations of the knowledge base. With the advent of the era of artificial intelligence, knowledge graphs have been heavily used, such as critical resources for intelligent applications such as intelligent question answering [ 5 ], web search [ 6 ], recommender system [ 7 ], and sentiment analysis [ 8 , 9 ]. Figure 1 is an example of a simple knowledge graph.…”

Section: Introductionmentioning

confidence: 99%

Convolutional Neural Network Knowledge Graph Link Prediction Model Based on Relational Memory

Shi

Jing

2023

Computational Intelligence and Neuroscience

View full text Add to dashboard Cite

A knowledge graph is a collection of fact triples, a semantic network composed of nodes and edges. Link prediction from knowledge graphs is used to reason about missing parts of triples. Common knowledge graph link prediction models include translation models, semantics matching models, and neural network models. However, the translation models and semantic matching models have relatively simple structures and poor expressiveness. The neural network model can easily ignore the overall structural characteristics of triples and cannot capture the links between entities and relations in low-dimensional space. In response to the above problems, we propose a knowledge graph embedding model based on a relational memory network and convolutional neural network (RMCNN). We encode triple embedding vectors using a relational memory network and decode using a convolutional neural network. First, we will obtain entity and relation vectors by encoding the latent dependencies between entities and relations and some critical information and keeping the translation properties of triples. Then, we compose a matrix of head entity encoding embedding vector, relation encoding embedding vector, and tail entity embedding encoding vector as the input of the convolutional neural network. Finally, we use a convolutional neural network as the decoder and a dimension conversion strategy to improve the information interaction capability of entities and relations in more dimensions. Experiments show that our model achieves significant progress and outperforms existing models and methods on several metrics.

show abstract

Multimodal Sentiment Analysis Based on Cross-Modal Attention and Gated Cyclic Hierarchical Fusion Networks

Cited by 8 publications

References 30 publications

Text Sentiment Analysis of Douban Film Short Comments Based on BERT-CNN-BiLSTM-Att Model

Text Sentiment Analysis of Douban Film Short Comments Based on BERT-CNN-BiLSTM-Att Model

Shaping high-performance wearable robots for human motor and sensory reconstruction and enhancement

Convolutional Neural Network Knowledge Graph Link Prediction Model Based on Relational Memory

Contact Info

Product

Resources

About