Co-Training for Cross-Lingual Sentiment Classification
By Xiaojun Wan. Annual Meeting of the Association for Computational Linguistics, 2009
546 citations recorded by the source corpus.
The lack of Chinese sentiment corpora limits the research progress on Chinese sentiment classification. However, there are many freely available English sentiment corpora on the Web. This paper focuses on the problem of cross-lingual sentiment classification, which leverages an available English corpus for Chinese sentiment classification by using the English corpus as training data. Machine translation services are used for eliminating the language gap between the training set and test set, and English features and Chinese features are considered as two independent views of the classification problem. We propose a cotraining approach to making use of unlabeled Chinese data. Experimental results show the effectiveness of the proposed approach, which can outperform the standard inductive classifiers and the transductive classifiers.
Subfields: Artificial Intelligence
View Paper: https://www.semanticscholar.org/paper/a3e39e183cf46a49320be761242c42e02e530e80
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Artificial Intelligence
Citations in the corpus
- Cross Language Text Classification via Subspace Co-regularized Multi-view Learning (2012, methodology): “A few multi-view learning meth-ods then have been applied on such multi-view data, including the co-training method (Wan, 2009) which is an instance of the standard co-training algorithm of (Blum & Mitchell, 1998), the multi-view majority voting method (Amini et al., 2009), and the multi-view…”
- A review of natural language processing techniques for opinion mining systems (2017, background): “Wang and Manning [34] analyzed the influences of features and datasets used in previous work and presented an SVM with näıve Bayes features.”
- Ranking Transfer Languages with Pragmatically-Motivated Features for Multilingual Sentiment Analysis (2020, methodology): “Traditional methods in CLSC often rely on machine translation systems (Wan, 2009) or bilin-”
- Attention-based LSTM Network for Cross-Lingual Sentiment Classification (2016, unknown): “Wan (2009) translated both the training data (English to Chinese) and the test data (Chinese to English) to train different models in both the source and target languages.”
- A survey on semi-supervised learning (2019, methodology): “Such an approach was first proposed by Yan andWang (2009), based on the sparse coding approach formulated for face recognition by Wright et al. (2009). The idea is to find, for each node xi , a coefficient vector a ∈ Rn denoting the contributions of all other nodes to the reconstruction ofxi .”
- Asymmetric Tri-training for Unsupervised Domain Adaptation (2017, methodology): “The generalization ability of co-training is theoretically ensured (Balcan et al., 2004; Dasgupta et al., 2001) under some assumptions and applied to various tasks (Wan, 2009; Levin et al., 2003).”
- Exploring Fine-tuning Techniques for Pre-trained Cross-lingual Models via Continual Learning (2020, background): “Cross-lingual models alleviate the need for obtaining annotated data in target languages (Bel et al., 2003; Wan, 2009), which copes with the data scarcity problem (Lample et al., 2018; Winata et al., 2020; Liu et al., 2020).”
- A survey of sentiment analysis in social media (2018, methodology): “Wan [133] proposed another co-training method which leverages an available English corpus for Chinese sentiment classification by using the English corpus as training data.”
- Cross-language web page classification via dual knowledge transfer using nonnegative matrix tri-factorization (2011, methodology): “1 Challenges in Cross-Language Web Page Classification and Our Motivations One of the most widely used strategy in cross-language Web page (text) classification is using language translation [15, 16, 18–21, 23].”
- Multilingual Deep Learning (2013, background): “To overcome this problem of resource scarcity, recently there as been a lot of interest in reusing resources from a resource fortunate language to develop NLP capabilities in a resource deprived language [25] [6] [14] [23] [17] [15] [13] [10].”
- Cross-Lingual Sentiment Classification with Bilingual Document Representation Learning (2016, methodology): “Wan (2009) translates both the training data (English to Chinese) and the test data (Chinese to English) to train different models in both the source and target languages.”
- A Survey of Opinion Mining and Sentiment Analysis (2012, background): “Machine translation is typically used [3, 8, 27, 47, 101].”