Sentiment classification of Chinese Weibo based on extended sentiment dictionary and organisational structure of comments

Zhang, Wei; Liu, Wenjuan; Zhu, Guangli; Zhang, Shunxiang; Hsieh, Meng-Yen

doi:10.1080/09540091.2021.2006146

Cited by 16 publications

(4 citation statements)

References 28 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In the course of our investigation into sentiment analysis [2] , [3] , [4] , we have deliberately directed our attention towards the pivotal role that datasets play in shaping the outcomes of sentiment analysis models. As our research unfolded, we observed a discernible pattern in the structure of topics and reviews prevalent in social media contexts [5] . This observation prompted the conceptualization and implementation of a dataset aligning with similar structures in 2022.…”

Section: Introductionmentioning

confidence: 91%

A structured sentiment analysis dataset based on public comments from various domains

Wei,

Zhang

2024

Data in Brief

Self Cite

View full text Add to dashboard Cite

Section: Introductionmentioning

confidence: 91%

A structured sentiment analysis dataset based on public comments from various domains

Wei,

Zhang

2024

Data in Brief

Self Cite

View full text Add to dashboard Cite

“…Sentiment classification, a way to identify the subjective information for the given context, has a wide range of applications across various industries and domains such as social media monitoring [38], customer feedback analysis [39], and financial forecasting [40]. Traditional supervised machine learning approaches have been widely investigated in this research area.…”

Section: Sentiment Classificationmentioning

confidence: 99%

Heterogeneous Student Knowledge Distillation From BERT Using a Lightweight Ensemble Framework

Lin,

Tsai,

Jwo

et al. 2024

IEEE Access

View full text Add to dashboard Cite

Deep learning models have demonstrated their effectiveness in capturing complex relationships between input features and target outputs across many different application domains. These models, however, often come with considerable memory and computational demands, posing challenges for deployment on resource-constrained edge devices. Knowledge distillation is a prominent technique for transferring the expertise from an advanced yet heavy teacher model to a more efficient leaner student model. As ensemble methods have exhibited notable enhancements in model generalization and have achieved state-of-the-art performance in various machine learning tasks, we adopt ensemble techniques to perform knowledge distillation from BERT using multiple lightweight student models. Our approach applies lean architectural paradigms of spatial and sequential networks including LSTM, CNN and their fusion to perform data processing from distinct perspectives. Instead of using contextual word representations which require more space in natural language processing applications, we take advantage of a single static pre-trained and lowdimensional word embedding space to be shared among student models. Empirical studies are conducted on the sentiment classification problem and our model outperforms not only other existing techniques but also the teacher model.INDEX TERMS Knowledge distillation, ensemble methods, BERT, LSTM, CNN, contextual word representations, pre-trained and low-dimensional word embedding space, sentiment classification problem.

show abstract

“…used the document topic generation model LDA to perform semantic downscaling and deep semantic feature extraction of microblogs. They used the ensemble classification method AdaBoost to identify the sentiment tendencies of comments below a microblog and validated its good performance in classifying the sentiment tendencies of Weibo users [ 1 ]. analyzed the accuracy of sentiment analysis models that combine the distributed word vector training method Word2vec and different classifiers in classifying the sentiments of online comments the public posted below the government affairs microblogs.…”

Section: Related Workmentioning

confidence: 99%

“…Existing research on sentiment analysis on Chinese government affairs microblogs primarily focuses on the sentiment classification of small-scale comments below a government affairs microblog [ 1 ]. The predictions involved in the analysis process typically use models based on traditional machine learning [ 2 ].…”

Section: Introductionmentioning

confidence: 99%

Sentiment analysis and prediction model based on Chinese government affairs microblogs

Meng

Shi

2023

Heliyon

View full text Add to dashboard Cite

Sentiment classification of Chinese Weibo based on extended sentiment dictionary and organisational structure of comments

Cited by 16 publications

References 28 publications

A structured sentiment analysis dataset based on public comments from various domains

A structured sentiment analysis dataset based on public comments from various domains

Heterogeneous Student Knowledge Distillation From BERT Using a Lightweight Ensemble Framework

Sentiment analysis and prediction model based on Chinese government affairs microblogs

Contact Info

Product

Resources

About