Leveraging bilingual-view parallel translation for code-switched emotion detection with adversarial dual-channel encoder

Zhu, Xun; Lou, Yinxia; Deng, Hongtao; Ji, Donghong

doi:10.1016/j.knosys.2021.107436

Cited by 10 publications

(5 citation statements)

References 47 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Zhu X. et al [7] translated the transcoding text into two languages, proposed an adversarial dual channel encoder architecture, in which two dedicated encoders took the parallel text of two languages as input respectively. Private coder and shared coder work together to effectively retrieve features from both monolingual and bilingual perspectives under confrontation training.…”

Section: Literature Reviewmentioning

confidence: 99%

A Generative Adversarial Networks Based Approach for Literary Translation

2023

Teh. vjesn.

View full text Add to dashboard Cite

This study aims to solve the problem of mistranslation due to the fact that literary intelligent translation only stays at the stage of text description and elaboration and lacks relevant facts. Therefore, this paper puts forward an improvement method of literary intelligent translation text based on generation confrontation network. First, an adaptive literary intelligent translation mode is designed under the generation confrontation network, and then the data of literary intelligent translation text improvement is preprocessed, and the data mining of text improvement quality evaluation is carried out. According to the mining results, a literary intelligent translation text improvement quality evaluation model is constructed to evaluate the quality of literary intelligent translation text improvement. According to the quality results, this paper constructs the improvement model of literary intelligent translation text, designs the improvement process, and completes the research on the improvement method of literary intelligent translation text that generates confrontation network. The experimental results show that this method has better detection effect of mistranslation features, better stability of the improved method, accurate and reliable results, and can improve the literary literacy of students and teachers.

show abstract

Section: Literature Reviewmentioning

confidence: 99%

A Generative Adversarial Networks Based Approach for Literary Translation

2023

Teh. vjesn.

View full text Add to dashboard Cite

show abstract

“…Emotion and code-mixing. Existing research on emotion analysis for code-mixed language primarily focuses on standalone social media texts (Sasidhar et al, 2020;Ilyas et al, 2023;Wadhawan and Aggarwal, 2021) and reviews (Suciati and Budi, 2020;Zhu et al, 2022). While aspects such as sarcasm (Kumar et al, 2022a,b), humour (Bedi et al, 2023), and offense (Madhu et al, 2023) have been explored within code-mixed conversations, emotion analysis remains largely an uncharted territory with no relevant literature available, to the best of our knowledge.…”

Section: Related Workmentioning

confidence: 99%

From Multilingual Complexity to Emotional Clarity: Leveraging Commonsense to Unveil Emotions in Code-Mixed Dialogues

Kumar,

Akhtar

et al. 2023

Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing

View full text Add to dashboard Cite

Understanding emotions during conversation is a fundamental aspect of human communication, driving NLP research for Emotion Recognition in Conversation (ERC). While considerable research has focused on discerning emotions of individual speakers in monolingual dialogues, understanding the emotional dynamics in codemixed conversations has received relatively less attention. This motivates our undertaking of ERC for code-mixed conversations in this study. Recognizing that emotional intelligence encompasses a comprehension of worldly knowledge, we propose an innovative approach that integrates commonsense information with dialogue context to facilitate a deeper understanding of emotions. To achieve this, we devise an efficient pipeline that extracts relevant commonsense from existing knowledge graphs based on the code-mixed input. Subsequently, we develop an advanced fusion technique that seamlessly combines the acquired commonsense information with the dialogue representation obtained from a dedicated dialogue understanding module. Our comprehensive experimentation showcases the substantial performance improvement obtained through the systematic incorporation of commonsense in ERC. Both quantitative assessments and qualitative analyses further corroborate the validity of our hypothesis, reaffirming the pivotal role of commonsense integration in enhancing ERC.

show abstract

“…The work carried out in [35] used machine learning techniques to find sentiments within images, and [36] built a novel image prediction algorithm while [37] presented techniques to analyze YouTube comments. [38] Improved code-switched emotion detection model and proposed dual-channel encoder. Construction of an emotion recognition system based on EEG is done in [39].…”

Section: B Machine Learning Based Approachmentioning

confidence: 99%

Emotion Quantification Using Variational Quantum State Fidelity Estimation

et al. 2022

View full text Add to dashboard Cite

Sentiment analysis has been instrumental in developing artificial intelligence when applied to various domains. However, most sentiments and emotions are temporal and often exist in a complex manner. Several emotions can be experienced at the same time. Instead of recognizing only categorical information about emotions, there is a need to understand and quantify the intensity of emotions. The proposed research intends to investigate a quantuminspired approach for quantifying emotional intensities in runtime. The inspiration comes from manifesting human cognition and decision-making capabilities, which may adopt a brief explanation through quantum theory. Quantum state fidelity was used to characterize states and estimate emotion intensities rendered by subjects from the Amsterdam Dynamic Facial Expression Set (ADFES) dataset. The Quantum variational classifier technique was used to perform this experiment on the IBM Quantum Experience platform. The proposed method successfully quantifies the intensities of joy, sadness, contempt, anger, surprise, and fear emotions of labelled subjects from the ADFES dataset.

show abstract

Leveraging bilingual-view parallel translation for code-switched emotion detection with adversarial dual-channel encoder

Cited by 10 publications

References 47 publications

A Generative Adversarial Networks Based Approach for Literary Translation

A Generative Adversarial Networks Based Approach for Literary Translation

From Multilingual Complexity to Emotional Clarity: Leveraging Commonsense to Unveil Emotions in Code-Mixed Dialogues

Emotion Quantification Using Variational Quantum State Fidelity Estimation

Contact Info

Product

Resources

About