Dependency Tree-based Sentiment Classification using CRFs with Hidden Variables

By Tetsuji Nakagawa, Kentaro Inui, Sadao Kurohashi. 2010

383 citations recorded by the source corpus.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Natural Language Processing

Citations in the corpus

Recursive Deep Models for Semantic Compositionality Over a Sentiment Treebank (2013, background): “Several works have explored sentiment compositionality through careful engineering of features or polarity shifting rules on syntactic structures (Polanyi and Zaenen, 2006; Moilanen and Pulman, 2007; Rentoumi et al., 2010; Nakagawa et al., 2010).”
Convolutional Neural Networks for Sentence Classification (2014, methodology): “CRF-PR : Conditional Random Fields with Posterior Regularization (Yang and Cardie, 2014).”
Discriminative Neural Sentence Modeling by Tree-Based Convolution (2015, background): “Feature engineering—for example, n-gram features (Cui et al., 2006), dependency subtree features (Nakagawa et al., 2010), or more dedicated ones (Silva et al., 2011)—can play an important role in modeling sentences.”
Semantic Compositionality through Recursive Matrix-Vector Spaces (2012, methodology): “Tree-CRF (Nakagawa et al., 2010) 77.”
Compositional Matrix-Space Models for Sentiment Analysis (2011, methodology): “For example, one can use the output of a dependency parser to guide the order of composition, similar to Nakagawa et al. (2010).”
Grammatical structures for word-level sentiment detection (2012, methodology): “Other work (Nakagawa et al., 2010) is similar to ours in that it uses factor graph modeling over a dependency parse formalism, but it assumes that opinionated language is known a priori and focuses on polarity classification, while our work tackles the more fundamental problem of identifying the opi”
Incorporating Phrase-level Sentiment Analysis on Textual Reviews for Personalized Recommendation (2015, methodology): “…the Netflix grand prize in the year of 2007 through 2009 [2], because the CFapproaches, especially those based on Matrix Factorization (MF) techniques [28, 8] on user-item numerical rating matrices, achieved important success in the task of rating prediction [12], and also exhibited great…”
Sequential Short-Text Classification with Recurrent and Convolutional Neural Networks (2016, methodology): “Many different approaches have been developed for short-text classiﬁcation, such as using Support Vector Machines (SVMs) with rule-based features (Silva et al., 2011), combining SVMs with naive Bayes (Wang and Manning, 2012), and building dependency trees with Conditional Random Fields (Nakagawa et ”
Explicit factor models for explainable recommendation based on phrase-level sentiment analysis (2014, background): “Review- [25, 42] and sentence-level [40, 22] analysis attempt to classify the sentiment of a whole review or sentence to one of the predefined sentiment polarities, including positive, negative and sometimes neutral, while phrase-level analysis [41, 19, 6] attempts to extract explicit features of th”
Beneath the Tip of the Iceberg: Current Challenges and New Directions in Sentiment Analysis Research (2020, neutral): “, 2010), syntactic dependencies (Nakagawa et al., 2010; Hutto & Gilbert, 2014), amongst others.”
Fracking Sarcasm using Neural Network (2016, methodology): “(Nakagawa et al., 2010) describes a Tree-CRF classiﬁer which uses a data-driven dependency parser, maltparser 2 , to obtain a parse tree for a sentence, and whose composition function uses the head-modiﬁer relations of the parse tree.”
The Role of Syntax in Vector Space Models of Compositional Semantics (2013, methodology): “We apply the same pre-processing as (Nakagawa et al., 2010) and (Socher et al., 2011b) by using an additional sentiment lexicon (Wilson et al., 2005) during the model training for this experiment.”
Recursive deep learning for natural language processing and computer vision (2014, background): “have explored sentiment compositionality through careful engineering of features or polarity shifting rules on syntactic structures (Polanyi and Zaenen, 2006; Nakagawa et al., 2010).”
Baselines and Bigrams: Simple, Good Sentiment and Topic Classification (2012, methodology): “(Nakagawa et al., 2010) used a SVM with secondorder polynomial kernel and additional features.”
Sentiment Classification by Leveraging the Shared Knowledge from a Sequence of Domains (2019, methodology): “Earlier techniques use hand-crafted features and external resources [21,23,38,43,46], which involve a great deal of human efforts.”