Sentence representation with manifold learning for biomedical texts

Zhao, Di; Wang, Jian; Lin, Hongfei; Chu, Yonghe; Wang, Yan; Zhang, Yijia; Yang, Zhihao

doi:10.1016/j.knosys.2021.106869

Cited by 22 publications

(12 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Table 3 shows the accuracy of BRCNN in the English tense collocation proposed in this study. To objectively evaluate the results of temporal collocation, HSASRL [23], SRML [24], and Usr-mtl [25] are introduced as comparison models. It can be seen from the results in Table 3 that the accuracy of BRCNN is higher than HSASRL and Usr-mtl.…”

Section: Results and Analysismentioning

confidence: 99%

Intelligent English Tense Collocation and Evaluation Based on Deep Reinforcement Learning

Ding

Tian-hua

2022

Mobile Information Systems

View full text Add to dashboard Cite

The representation of time in sentences is the key problem for tense collocation. Based on the relationship expression among regions in Allen’s interval algebra theory, we propose a vector representation method, i.e., relationship vector, and several operations are defined based on the relationship vector for temporal reasoning in this work. This method transforms the original matrix representation into vector representation, which reduces the amount of computation of temporal reasoning. In addition, we propose a temporal classification and collocation method based on deep learning and deep reinforcement learning. This method uses a bidirectional cyclic neural network and a convolutional neural network for text expression and achieves temporal word classification and temporal collocation based on the deep reinforcement learning model. In the experiments, the proposed method obtains the average accuracy of 92.17% in five datasets, i.e., MPQA, CR, MR, Subj, and TREC, which proves its effectiveness in tense collocation.

show abstract

Section: Results and Analysismentioning

confidence: 99%

Intelligent English Tense Collocation and Evaluation Based on Deep Reinforcement Learning

Ding

Tian-hua

2022

Mobile Information Systems

View full text Add to dashboard Cite

show abstract

“…Good results are obtained in multiple datasets. Zhao et al [19] use the attention mechanism to obtain semantic representation at different levels in sentences, which can more accurately express the emotion reflected in the text and make the sentence representation more comprehensive. Most of the neural network training models remain in the training of vocabulary targets, and there are few studies on the whole sentence as the training target.…”

Section: Related Workmentioning

confidence: 99%

CharAs-CBert: Character Assist Construction-Bert Sentence Representation Improving Sentiment Classification

Chen

Peng

Song

2022

Sensors

View full text Add to dashboard Cite

In the process of semantic capture, traditional sentence representation methods tend to lose a lot of global and contextual semantics and ignore the internal structure information of words in sentences. To address these limitations, we propose a sentence representation method for character-assisted construction-Bert (CharAs-CBert) to improve the accuracy of sentiment text classification. First, based on the construction, a more effective construction vector is generated to distinguish the basic morphology of the sentence and reduce the ambiguity of the same word in different sentences. At the same time, it aims to strengthen the representation of salient words and effectively capture contextual semantics. Second, character feature vectors are introduced to explore the internal structure information of sentences and improve the representation ability of local and global semantics. Then, to make the sentence representation have better stability and robustness, character information, word information, and construction vectors are combined and used together for sentence representation. Finally, the evaluation and verification are carried out on various open-source baseline data such as ACL-14 and SemEval 2014 to demonstrate the validity and reliability of sentence representation, namely, the F1 and ACC are 87.54% and 92.88% on ACL14, respectively.

show abstract

“…That is, sentence representations are gradually combined with neural networks from simple word embedding models. For example, Zhao et al [17] considered the powerful feature extraction capability of a convolutional neural network (CNN), which captures feature information at different levels of text and performs simple splicing to generate sentence embedding representations. Although CNN can obtain the local information flow of the sentence well, it is easy to lose the global semantics of the input sentence, resulting in problems such as poor sentence representation.…”

Section: Relate Workmentioning

confidence: 99%

MAI-CBert: Multidirectional Attention Interaction Construction-Bert Sentiment Sentence Representation

2022

View full text Add to dashboard Cite

Considering that when the traditional network structure is connected with the pre-trained language model for sentence embedding representation, it relies on simple word weights to generate sentence vectors, which easily ignores the global and contextual semantic details of sentences, reducing the accuracy of sentence representation. To address these issues, we develop a Multidirectional Attention Interaction Construction-Bert Sentence Representation Framework (MAI-CBert). First, the model uses deep attention to transform the input emotional sentences to reduce misunderstanding caused by the unbalanced distribution of word weights in the sentences; at the same time, it applies horizontal and vertical attention to emotional constructions, which are generated from two different directions. Construction vectors, aim to focus on words with salient features in sentences and establish close dependencies. Second, the dynamic interaction strategy is adopted to realize the interaction between attention in different directions, so that the information flow forms a complementary relationship, resulting in more effective sentence vectors. Notably, the loss function is refactored to improve the representation robustness of the model to avoid catastrophic forgetting. Experimental results on the SemEval-14 and ACL-14 baseline datasets demonstrate that the MAI-CBert sentence representation framework is robust and competitive.

show abstract

Sentence representation with manifold learning for biomedical texts

Cited by 22 publications

References 26 publications

Intelligent English Tense Collocation and Evaluation Based on Deep Reinforcement Learning

Intelligent English Tense Collocation and Evaluation Based on Deep Reinforcement Learning

CharAs-CBert: Character Assist Construction-Bert Sentence Representation Improving Sentiment Classification

MAI-CBert: Multidirectional Attention Interaction Construction-Bert Sentiment Sentence Representation

Contact Info

Product

Resources

About