Back

Deep learning for sentiment analysis: A survey

Lei ZhangShuai WangBing Liu

article2018WIREs Data Mining Knowl. Discov.1,956 citations

Categorizes deep learning architectures and representation methods across document, sentence, and aspect levels of sentiment analysis, providing a structured reference for building modern opinion mining systems.

Listen

Modern decision-making across commercial, political, and social sectors relies heavily on understanding public opinions, emotions, and attitudes. With the exponential growth of digital text from social media, online reviews, and forums, manual monitoring and synthesis of this data have become impractical. Automated sentiment analysis systems are critical for organizations seeking real-time, scalable insights. The article provides a comprehensive survey of modern deep learning architectures and evaluates their effectiveness across various sentiment analysis tasks compared to traditional statistical and rule-based methods.

To conduct this evaluation, the article synthesizes extensive research spanning deep neural network architectures—such as feedforward networks, convolutional neural networks, recurrent and recursive neural networks, memory networks, and attention mechanisms—applied across multiple analytical granularities. It evaluates these models on key tasks, including document-level, sentence-level, and fine-grained aspect-level sentiment classification, alongside opinion extraction, sarcasm detection, emotion analysis, and multimodal processing.

The findings show that deep learning models consistently achieve state-of-the-art accuracy by automatically learning complex, hierarchical feature representations directly from data, eliminating the need for manual feature engineering. Word and document embeddings significantly outperform traditional bag-of-words representations by capturing semantic nuance and context. Across specific architectures, recurrent models with attention mechanisms and memory networks provide superior performance for fine-grained, aspect-based sentiment tasks by accurately linking opinions to specific target entities. Furthermore, deep learning techniques adapt effectively to complex scenarios, including cross-lingual analysis in resource-poor languages and multimodal settings integrating text, audio, and visual signals.

These findings indicate that adopting deep learning pipelines can substantially reduce development costs and labor traditionally associated with manual linguistic rule creation while improving analytical accuracy and real-time responsiveness. Organizations can capture more nuanced customer insights and brand perceptions, reducing the risk of misinterpreting complex expressions such as sarcasm or multi-aspect feedback.

Organizations deploying opinion mining should prioritize hybrid, attention-based deep learning frameworks and incorporate pre-trained, sentiment-specific word embeddings tailored to their operational domain. Before full-scale deployment, engineering teams should implement pilot evaluations to validate model performance across domain shifts and multilingual contexts. Continued monitoring is recommended, as deep neural models require substantial computational resources and large labeled datasets, and their performance can vary depending on task complexity and data quality.

LinkedInUniversity of Illinois Chicagodeep learningsentiment analysisconvolutional neural networksdeep learningneural sentiment analysisdocument-level sentiment classification
arXiv: 1801.07883
  • Paper: Deep contextualized word representations, Matthew E. Peters et al. (2018). It extends neural representations from static word vectors to deep contextualized embeddings (ELMo), significantly advancing state-of-the-art sentiment analysis pipelines.
  • Paper: Universal Sentence Encoder, Daniel Cer et al. (2018). It generalizes sentence-level modeling into universal pre-trained encoders using Transformer and DAN architectures to enhance transfer performance on sentiment benchmarks.
  • Paper: Graph Convolutional Networks for Text Classification, Liang Yao et al. (2018). It extends deep text classification beyond sequential neural nets to graph convolutional networks that model corpus-wide word-document relationships for opinion and sentiment mining.
  • Paper: A Survey on Deep Transfer Learning, Chuanqi Tan et al. (2018). It provides a comprehensive taxonomy of deep transfer learning methods, formalizing domain adaptation paradigms crucial for cross-domain sentiment classification.
  • Paper: Ensemble deep learning: A review, M. A. Ganaie et al. (2021). It reviews ensemble deep learning frameworks that can be applied to mitigate variance and improve robustness over standalone sentiment classifiers.
Cover for Deep learning for sentiment analysis: A survey

Abstract

Deep learning has emerged as a powerful machine learning technique that learns multiple layers of representations or features of the data and produces state-of-the-art prediction results. Along with the success of deep learning in many other application domains, deep learning is also popularly used in sentiment analysis in recent years. This paper first gives an overview of deep learning and then provides a comprehensive survey of its current applications in sentiment analysis.

Topics

deep learningsentiment analysisconvolutional neural networks

Keywords

deep learningneural sentiment analysisdocument-level sentiment classificationsentence sentiment classificationaspect-based Twitter sentiment classificationtarget-dependent sentiment classificationcross-language sentiment classificationopinion expression extractiondeep memory networkdeep bidirectional RNN

Table of Contents

  • INTRODUCTION
  • NEURAL NETWORKS
  • Training algorithm : stochastic gradient descent via backpropagation
  • DEEP LEARNING
  • WORD EMBEDDING
  • AUTOENCODER AND DENOISING AUTOENCODER
  • CONVOLUTIONAL NEURAL NETWORK
  • RECURRENT NEURAL NETWORK
  • LSTM NETWORK
  • ATTENTION MECHANISM WITH RECURRENT NEURAL NETWORK
  • MEMORY NETWORK
  • RECURSIVE NEURAL NETWORK
  • SENTIMENT ANALYSIS TASKS
  • DOCUMENT LEVEL SENTIMENT CLASSIFICATION
  • SENTENCE LEVEL SENTIMENT CLASSIFICATION
  • ASPECT LEVEL SENTIMENT CLASSIFICATION
  • ASPECT EXTRACTION AND CATEGORIZATION
  • OPINION EXPRESSION EXTRACTION
  • SENTIMENT COMPOSITION
  • OPINION HOLDER EXTRACTION
  • TEMPORAL OPINION MINING
  • SENTIMENT ANALYSIS WITH WORD EMBEDDING
  • SARCASM ANALYSIS
  • EMOTION ANALYSIS
  • MULTIMODAL DATA FOR SENTIMENT ANALYSIS
  • RESOURCE-POOR LANGUAGE AND MULTILINGUAL SENTIMENT ANALYSIS
  • OTHER RELATED TASKS
  • CONCLUSION
  • Acknowledgments
  • References

Knowls

  1. Knowl 1 — Taxonomy and Granularity of Sentiment Analysis Tasks

    definition

    Sentiment analysis (opinion mining) is computationally categorized according to the granularity of the target text into three fundamental operational levels:

    1. Document-Level Sentiment Classification: Analyzes an entire opinionated document under the assumption that it conveys sentiment about a single primary entity. The objective is to assign a single overall polarity (positive vs. negative) or infer an ordinal rating score (e.g., a 1-to-5 star rating).

    2. Sentence-Level Sentiment Classification: Analyzes individual sentences where opinionated status cannot be presupposed. This requires either a pipeline of subjectivity classification (distinguishing subjective from objective statements) followed by polarity classification, or a unified three-class categorization (positive, negative, neutral).

    3. Aspect-Level Sentiment Analysis (ABSA): Provides fine-grained analysis by decoupling sentiment targets into specific entities and their individual attributes (aspects). ABSA consists of two interrelated subtasks: aspect extraction (identifying target terms or phrases referring to entity attributes) and aspect sentiment classification (determining the specific sentiment polarity directed at each extracted aspect).

    Additional specialized subtasks include opinion holder extraction (identifying source entities), opinion expression extraction, sentiment compositionality modeling, emotion classification, sarcasm detection, and multimodal sentiment evaluation.

  2. Knowl 2 — Categorization Framework for Document-Level Deep Sentiment Classification

    data/table

    Deep learning approaches to document-level sentiment classification are systematically categorized across four core technical axes: text representation strategy, primary neural architecture, presence and type of attention mechanism, and integration of auxiliary metadata (such as user, product, or cross-domain features).

    Research Work Document/Text Representation Neural Network Model Attention Mechanism Joint Modelling with Sentiment
    Moraes et al. BoW ANN (Artificial Neural Network) No -
    Le and Mikolov Paragraph Vector Paragraph Vector No -
    Glorot et al. BoW to dense vector SDA (Stacked Denoising Autoencoder) No Unsupervised target domain representation
    Zhai and Zhang BoW to dense vector DAE (Denoising Autoencoder) No -
    Johnson and Zhang BoW to dense vector BoW-CNN and Seq-CNN No -
    Tang et al. Word embeddings to dense vector CNN/LSTM + GRU No -
    Tang et al. Word embeddings to dense vector UPNN (User Product Neural Network) based on CNN No User and product information
    Chen et al. Word embeddings to dense vector UPA (User Product Attention) based on LSTM Yes User and product information
    Dou Word embeddings to dense vector Memory Network Yes User and product information
    Xu et al. Word embeddings to dense vector Cached LSTM No -
    Yang et al. Word embeddings to dense vector GRU sequence encoder Hierarchical attention -
    Yin et al. Word embeddings to dense vector Input encoder and LSTM Hierarchical attention Aspect/target information
    Zhou et al. Word embeddings to dense vector LSTM Hierarchical attention Cross-lingual information
    Li et al. Word embeddings to dense vector Memory Network Yes Cross-domain information

    The taxonomy illustrates the evolution from standard bag-of-words (BoW) transformations through autoencoders to dense vector compositions via CNN/LSTM/GRU networks, culminating in hierarchical attention networks and memory networks incorporating auxiliary user and product contexts.

  3. Knowl 3 — Long Short-Term Memory (LSTM) Recurrent Architecture for Sentiment Modeling

    model/method

    The Long Short-Term Memory (LSTM) network resolves the vanishing and exploding gradient problems of standard recurrent neural networks when capturing long-range linguistic dependencies. At step tt, given input token vector xtx_t, previous hidden state ht−1h_{t-1}, and previous cell state Ct−1C_{t-1}, the module computes state transitions via affine transformations parameterized by weight matrices W,UW, U and bias vectors using non-linear activation functions σ\sigma (sigmoid) and tanh⁡\tanh:

    Forget Gate: ft=σ(Wfxt+Ufht−1)\text{Forget Gate: } f_t = \sigma(W^f x_t + U^f h_{t-1})

    Input Gate: it=σ(Wixt+Uiht−1)\text{Input Gate: } i_t = \sigma(W^i x_t + U^i h_{t-1})

    Candidate State: C~t=tanh⁡(Wnxt+Unht−1)\text{Candidate State: } \tilde{C}_t = \tanh(W^n x_t + U^n h_{t-1})

    Cell State Update: Ct=ft⊙Ct−1+it⊙C~t\text{Cell State Update: } C_t = f_t \odot C_{t-1} + i_t \odot \tilde{C}_t

    Output Gate: ot=σ(Woxt+Uoht−1)\text{Output Gate: } o_t = \sigma(W^o x_t + U^o h_{t-1})

    Hidden State: ht=ot⊙tanh⁡(Ct)\text{Hidden State: } h_t = o_t \odot \tanh(C_t)

    where ⊙\odot represents element-wise multiplication. The forget gate ft∈[0,1]df_t \in [0, 1]^d provides an explicit linear highway to delete or retain historical memory, preserving gradient flow through time during backpropagation.

  4. Knowl 4 — Deep Learning Paradigms for Aspect-Level Sentiment Classification

    model/method

    Aspect-level sentiment classification determines the sentiment polarity expressed toward a specific target aspect within a sentence containing potentially mixed sentiments across multiple targets. Neural approaches address this challenge through three primary paradigms:

    1. Target-Dependent Recurrent Models (TD-LSTM, TC-LSTM): Extend recurrent networks by conditioning sentence representations directly on target word embeddings. TD-LSTM splits the sentence at the target position using two LSTMs (left context and right context), while TC-LSTM concatenates target embedding vectors with word embeddings at each recurrent time step.

    2. Target-Specific Attention Networks (IAN, Attention-BiLSTM): Employ attention mechanisms where target aspect vectors query sentence hidden states to generate context weights, isolating aspect-relevant modifiers (e.g., attending to "clear" for "screen" and ignoring "short" for "battery"). Interactive Attention Networks (IAN) employ two interdependent attention networks to model target descriptors and context sentences simultaneously.

    3. Deep Memory Networks (MemN2N, DyMemNN): Treat contextual words as an external memory bank, employing multi-hop recurrent attention layers over memory slices to resolve complex, long-range dependencies, with dyadic variants modeling explicit tensor or holographic compositions between aspect and context representations.

  5. Knowl 5 — Deep Neural Architectures for Aspect Term Extraction and Categorization

    model/method

    Automated aspect term extraction identifies opinion targets in text, while aspect categorization groups distinct aspect expressions into canonical categories without manual feature engineering. Dominant deep neural frameworks include:

    1. Neural Conditional Random Fields (LSTM-CRF, RNN-CRF): Bidirectional LSTMs encode contextual semantic features from continuous word embeddings, which are fed into a linear-chain CRF output layer to jointly decode sequence tags while modeling label transition dependencies.

    2. Coupled Multi-Layer Attention Models (CMLA): Employ multi-layer attention networks using GRU cells to perform co-extraction of aspects and opinion words, allowing iterative bi-directional information propagation between aspect targets and opinion expressions.

    3. Convolutional Token Classifiers: Deep multi-layer CNNs tag each word directly as aspect or non-aspect using stacked convolutional feature extractors and linguistic patterns.

    4. Unsupervised Attention Autoencoders & Metric Learning: Use attention mechanisms to emphasize aspect-related terms while suppressing background words during representation learning, and employ deep metric learning algorithms over phrase embeddings for subsequent clustering into aspect categories.

  6. Knowl 6 — Recursive Neural Network Architectures for Sentiment Compositionality

    model/method

    Sentiment compositionality states that the overall polarity of a phrase or sentence is determined hierarchically by syntactic structure and constituent word semantics. Recursive Neural Networks (RecNN) model this by mapping constituent tree parse nodes into a shared continuous semantic vector space in a bottom-up structure:

    1. Standard RecNN & Recursive Autoencoders (RAE): Recursively combine child vector representations (c1,c2)(c_1, c_2) into a parent representation p=f(W[c1;c2]+b)p = f(W [c_1; c_2] + b) up to the root node, which is evaluated with a softmax classifier.

    2. Matrix-Vector Recursive Neural Network (MV-RNN): Represents every word/phrase as both a vector capturing its meaning and a matrix capturing its operator effect on neighboring constituents, modeling modifier interactions (e.g., negation and intensification).

    3. Recursive Neural Tensor Network (RNTN): Replaces standard affine combinations with a 3D tensor compositional function p=f(cTV[1:d]c+Wc+b)p = f(c^T V^{[1:d]} c + W c + b) where c=[c1;c2]c = [c_1; c_2], capturing multiplicative feature interactions between constituent sub-phrases.

    4. Tag-Guided / Tag-Embedded RecNNs (TG-RNN, TE-RNN): Dynamically select composition functions based on the part-of-speech tags of phrases, or jointly learn POS-tag embeddings combined with word representations.

  7. Knowl 7 — Sentiment-Specific and Polarity-Preserving Word Embeddings

    model/method

    Standard unsupervised word embeddings (such as Word2Vec CBOW, Skip-Gram, and GloVe) learn representations purely from contextual co-occurrence. Consequently, antonyms and words with opposite sentiment polarities occurring in identical syntactic environments (e.g., "good" versus "bad") are assigned geometrically proximate vector representations, degrading sentiment classification accuracy.

    To resolve this limitation, sentiment-aware embedding algorithms inject polarity supervision:

    • Sentiment-Specific Word Embeddings (SSWE): Optimize a joint loss combining context-prediction objectives with sentence/document sentiment label classification, forcing opposite-polarity words into distinct regions of the embedding space.
    • Lexicon-Guided Vector Refinement: Existing dense embeddings are projected or regularized via optimization frameworks using sentiment polarity lexicons, pulling same-polarity words closer while increasing the distance between opposing sentiment words.
    • Bilingual Sentiment Word Embeddings (BSWE): Jointly constrain cross-lingual semantic representations using sentiment-annotated corpora from a source language and translation pairs to map sentiment semantics across languages without requiring massive sentence-aligned parallel corpora.

  8. Knowl 8 — Stochastic Gradient Descent via Backpropagation for Neural Sentiment Models

    algorithm

    Supervised training of feedforward and recurrent neural networks for sentiment classification optimizes weight matrices and bias vectors by minimizing the cross-entropy loss between softmax predictions and ground-truth sentiment labels using stochastic gradient descent and recursive backward chain-rule gradient propagation.

    Input: Training dataset D={(xi,yi)}i=1ND = \{ (x_i, y_i) \}_{i=1}^N of sample feature vectors xix_i and true sentiment labels yiy_i, learning rate η\eta
    Output: Trained neural network parameters (W,b)(W, b)
    Initialize weight matrices WW and bias vectors bb with random values
    repeat
        for each training example (xi,yi)∈D(x_i, y_i) \in D do
            Compute forward prediction probability pi=ForwardPass(N,xi)p_i = \text{ForwardPass}(N, x_i)
            Calculate gradient of cross-entropy loss L(pi,yi)\mathcal{L}(p_i, y_i) with respect to output layer weights wLw^L
            Compute parameter update ΔwL\Delta w^L
            for layer l=L−1l = L-1 down to 11 do
                Calculate error gradients with respect to hidden layer weights wlw^l using backpropagation chain rule
                Compute parameter update Δwl\Delta w^l
            end for
            Update parameter tensors: (W,b)←(W,b)−η(ΔW,Δb)(W, b) \leftarrow (W, b) - \eta (\Delta W, \Delta b)
        end for
    until stopping criteria are met or all training examples are classified correctly
    return Trained neural network model

  9. Knowl 9 — Multimodal Data Fusion Models for Sentiment and Emotion Recognition

    model/method

    Multimodal sentiment analysis incorporates auxiliary non-textual modalities—primarily visual features (facial expressions, image scenery) and acoustic features (pitch, speech prosody) alongside textual data. Core neural fusion strategies include:

    1. Intermediate Layer Concatenation & Multiple Kernel Learning (MKL): CNN feature representations extracted from text and acoustic/visual signals are concatenated into a joint representation vector and classified via Support Vector Machines with multiple kernel learning.

    2. Deep Coupled Adjective-Noun Networks (DCAN): For visual sentiment, adjective and noun image descriptions act as dual weakly supervised signals, learning intermediate visual representations before concatenation for classification.

    3. Tensor Fusion Networks (TFN): Explicitly model intra-modality dynamics within individual channels and inter-modality dynamics across text, visual, and audio streams by computing outer product tensors across unimodal representations.

    4. Cognition-Based Attention (CBA): Uses human gaze and eye-tracking patterns as cognitive supervisory signals to weight textual features during convolutional and recurrent sentence classification.

  10. Knowl 10 — Deep Learning Architectures for Sarcasm Detection and Emotion Analysis

    model/method

    Sarcasm detection and fine-grained emotion analysis extend standard polarity classification to figurative and psychological language phenomena:

    1. Sarcasm Detection Networks: Sarcasm manifests through semantic dissonance between expressed words and intended meaning. Hybrid deep learning architectures combine CNNs (capturing coarse-grained local semantic indicators) with bidirectional LSTMs/GRUs to model psychological and long-range contextual discourse. Tweet-level sarcasm models leverage bidirectional GRUs over immediate text paired with pooling networks over historical author tweets to capture contextual contrast, as well as distant supervision pretraining on emoji occurrences.

    2. Emotion Analysis and Cause Extraction: Goes beyond positive/negative valence to classify primary emotion categories (joy, anger, sadness, fear, surprise). Models employ bilingual attention networks for code-switched posts, Emotional Chatting Machines using GRU-based internal and external memory to track emotional states during dialogue generation, and deep memory networks formulated as question-answering systems to extract text rationales representing emotion causes.

Coverage note — Omitted brief citations of domain-specific single papers (e.g., basic financial volatility prediction and standard stance classification variants) as they represent standard supervised applications covered under the general classification paradigms.

References

  1. 1.Liu B. Sentiment analysis: mining opinions, sentiments, and emotions. The Cambridge University Press, 2015.
  2. 2.Liu B. Sentiment analysis and opinion mining (introduction and survey), Morgan & Claypool, May 2012.
  3. 3.Pang B and Lee L. Opinion mining and sentiment analysis. Foundations and Trends in Information Retrieval, 2008. 2(1–2): pp. 1–135.
  4. 4.Goodfellow I, Bengio Y, Courville A. Deep learning. The MIT Press. 2016.
  5. 5.Glorot X, Bordes A, Bengio Y. Deep sparse rectifier neural networks. In Proceedings of the International Conference on Artificial Intelligence and Statistics (AISTATS 2011), 2011.
  6. 6.Rumelhart D.E, Hinton G.E, Williams R.J. Learning representations by back-propagating errors. Cognitive modelling, 1988.
  7. 7.Collobert R, Weston J, Bottou L, Karlen M, Kavukcuoglu K, and Kuksa P. Natural language processing (almost) from scratch. Journal of Machine Learning Research, 2011.
  8. 8.Goldberg Y. A primer on neural network models for natural language processing. Journal of Artificial Intelligence Research, 2016.
  9. 9.Bengio Y, Courville A, Vincent P. Representation learning: a review and new perspectives. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2013.
  10. 10.Lee H, Grosse R, Ranganath R, and Ng A.Y. Convolutional deep belief networks for scalable unsupervised learning of hierarchical representations. In Proceedings of the International Conference on Machine Learning (ICML 2009), 2009.
  11. 11.Bengio Y, Ducharme R, Vincent P, and Jauvin C. A neural probabilistic language model. Journal of Machine Learning Research, 2003.
  12. 12.Morin F, Bengio Y. Hierarchical probabilistic neural network language model. In Proceedings of the International Workshop on Artificial Intelligence and Statistics, 2005.
  13. 13.Mikolov T, Chen K, Corrado G, and Dean J. Efficient estimation of word representations in vector space. In Proceedings of International Conference on Learning Representations (ICLR 2013), 2013.
  14. 14.Mikolov T, Sutskever I, Chen K, Corrado G, and Dean J. Distributed representations of words and phrases and their compositionality. In Proceedings of the Annual Conference on Advances in Neural Information Processing Systems (NIPS 2013), 2013.
  15. 15.Mnih A, Kavukcuoglu K. Learning word embeddings efficiently with noise-contrastive estimation. In Proceedings of the Annual Conference on Advances in Neural Information Processing Systems (NIPS 2013), 2013.
  16. 16.Huang E.H, Socher R, Manning C.D. and Ng A.Y. Improving word representations via global context and multiple word prototypes. In Proceedings of the Annual Meeting of the Association for Computational Linguistics (ACL 2012), 2012.
  17. 17.Pennington J, Socher R, Manning C.D. GloVe: global vectors for word representation. In Proceedings of the Conference on Empirical Methods on Natural Language Processing (EMNLP 2014), 2014.
  18. 18.Bengio Y, Lamblin P, Popovici D, and Larochelle H. Greedy layer-wise training of deep networks. In Proceedings of the Annual Conference on Advances in Neural Information Processing Systems (NIPS 2006), 2006.
  19. 19.Hinton G.E, Salakhutdinov R.R. Reducing the dimensionality of data with neural networks. Science, July 2006.
  20. 20.Vincent P, Larochelle H, Bengio Y, and Manzagol P-A. Extracting and composing robust features with denoising autoencoders. In Proceedings of the International Conference on Machine Learning (ICML 2008), 2008.
  21. 21.Sermanet P, LeCun Y. Traffic sign recognition with multi-scale convolutional networks. In Proceedings of the International Joint Conference on Neural Networks (IJCNN 2011), 2011.
  22. 22.Elman J.L. Finding structure in time. Cognitive Science, 1990.
  23. 23.Bengio Y, Simard P, Frasconi P. Learning long-term dependencies with gradient descent is difficult. IEEE Transactions on Neural Networks, 1994.
  24. 24.Schuster M, Paliwal K.K. Bidirectional recurrent neural networks. IEEE Transactions on Signal Processing, 1997.
  25. 25.Hochreiter S, Schmidhuber J. Long short-term memory. Neural Computation, 9(8): 1735-1780, 1997.
  26. 26.Tai K.S, Socher R, Manning C. D. Improved semantic representations from tree-structured long short-term memory networks. In Proceedings of the Annual Meeting of the Association for Computational Linguistics (ACL 2015), 2015.
  27. 27.Cho K, Bahdanau D, Bougares F, Schwenk H and Bengio Y. Learning phrase representations using RNN encoder-decoder for statistical machine translation. In Proceedings of the Conference on Empirical Methods in Natural Language Processing (EMNLP 2014), 2014.
  28. 28.Chung J, Gulcehre C, Cho K, Bengio Y. Empirical evaluation of gated recurrent neural networks on sequence modelling. arXiv preprint arXiv:1412.3555, 2014.
  29. 29.Bahdanau D, Cho K, Bengio Y. Neural machine translation by jointly learning to align and translate. arXiv preprint arXiv:1409.0473, 2014.
  30. 30.Weston J, Chopra S, Bordes A. Memory networks. arXiv preprint arXiv:1410.3916. 2014.
  31. 31.Sukhbaatar S, Weston J, Fergus R. End-to-end memory networks. In Proceedings of the 29th Conference on Neural Information Processing Systems (NIPS 2015), 2015.
  32. 32.Graves A, Wayne G, Danihelka I. Neural Turing Machines. preprint arXiv:1410.5401. 2014.
  33. 33.Qian Q, Tian B, Huang M, Liu Y, Zhu X and Zhu X. Learning tag embeddings and tag-specific composition functions in the recursive neural network. In Proceedings of the Annual Meeting of the Association for Computational Linguistics (ACL 2015), 2015.
  34. 34.Moraes R, Valiati J.F, Neto W.P. Document-level sentiment classification: an empirical comparison between SVM and ANN. Expert Systems with Applications. 2013.
  35. 35.Le Q, Mikolov T. Distributed representations of sentences and documents. In Proceedings of the International Conference on Machine Learning (ICML 2014), 2014.
  36. 36.Glorot X, Bordes A, Bengio Y. Domain adaption for large-scale sentiment classification: a deep learning approach. In Proceedings of the International Conference on Machine Learning (ICML 2011), 2011.
  37. 37.Zhai S, Zhongfei (Mark) Zhang. Semisupervised autoencoder for sentiment analysis. In Proceedings of AAAI Conference on Artificial Intelligence (AAAI 2016), 2016.
  38. 38.Johnson R, Zhang T. Effective use of word order for text categorization with convolutional neural networks. In Proceedings of the Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (NAACL-HLT 2015), 2015.
  39. 39.Tang D, Qin B, Liu T. Document modelling with gated recurrent neural network for sentiment classification. In Proceedings of the Conference on Empirical Methods in Natural Language Processing (EMNLP 2015), 2015.
  40. 40.Tang D, Qin B, Liu T. Learning semantic representations of users and products for document level sentiment classification. In Proceedings of the Annual Meeting of the Association for Computational Linguistics (ACL 2015), 2015.
  41. 41.Chen H, Sun M, Tu C, Lin Y, and Liu Z. Neural sentiment classification with user and product attention. In Proceedings of the Conference on Empirical Methods in Natural Language Processing (EMNLP 2016), 2016.
  42. 42.Dou ZY. Capturing user and product Information for document level sentiment analysis with deep memory network. In Proceedings of the Conference on Empirical Methods on Natural Language Processing (EMNLP 2017), 2017.
  43. 43.Xu J, Chen D, Qiu X, and Huang X. Cached long short-term memory neural networks for document-level sentiment classification. In Proceedings of the Conference on Empirical Methods in Natural Language Processing (EMNLP 2016), 2016.
  44. 44.Yang Z, Yang D, Dyer C, He X, Smola AJ, and Hovy EH. Hierarchical attention networks for document classification. In Proceedings of the Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (NAACL-HLT 2016), 2016.
  45. 45.Yin Y, Song Y, Zhang M. Document-level multi-aspect sentiment classification as machine comprehension. In Proceedings of the Conference on Empirical Methods in Natural Language Processing (EMNLP 2017), 2017.
  46. 46.Zhou X, Wan X, Xiao J. Attention-based LSTM network for cross-lingual sentiment classification. In Proceedings of the Conference on Empirical Methods in Natural Language Processing (EMNLP 2016), 2016.
  47. 47.Li Z, Zhang Y, Wei Y, Wu Y, and Yang Q. End-to-end adversarial memory network for cross-domain sentiment classification. In Proceedings of the International Joint Conference on Artificial Intelligence (IJCAI 2017), 2017.
  48. 48.Wiebe J, Bruce R, and O’Hara T. Development and use of a gold standard data set for subjectivity classifications. In Proceedings of the Annual Meeting of the Association for Computational Linguistics (ACL 1999), 1999.
  49. 49.Socher R, Pennington J, Huang E.H, Ng A.Y, and Manning C.D. Semi-supervised recursive autoencoders for predicting sentiment distributions. In Proceedings of the Conference on Empirical Methods in Natural Language Processing (EMNLP 2011), 2011.
  50. 50.Socher R, Huval B, Manning C.D, and Ng A.Y. Semantic compositionality through recursive matrix-vector spaces. In Proceedings of the Conference on Empirical Methods on Natural Language Processing (EMNLP 2012), 2012.
  51. 51.Socher R, Perelygin A, Wu J. Y, Chuang J, Manning C.D, Ng A. Y, and Potts C. Recursive deep models for semantic compositionality over a sentiment treebank. In Proceedings of the Conference on Empirical Methods on Natural Language Processing (EMNLP 2013), 2013.
  52. 52.Kalchbrenner N, Grefenstette E, Blunsom P. A convolutional neural network for modelling sentences. In Proceedings of the Annual Meeting of the Association for Computational Linguistics (ACL 2014), 2014.
  53. 53.Kim Y. Convolutional neural networks for sentence classification. In Proceedings of the Annual Meeting of the Association for Computational Linguistics (ACL 2014), 2014.
  54. 54.dos Santos, C. N., Gatti M. Deep convolutional neural networks for sentiment analysis for short texts. In Proceedings of the International Conference on Computational Linguistics (COLING 2014), 2014.
  55. 55.Wang X, Liu Y, Sun C, Wang B, and Wang X. Predicting polarities of tweets by composing word embeddings with long short-term memory. In Proceedings of the Annual Meeting of the Association for Computational Linguistics (ACL 2015), 2015.
  56. 56.Graves A, Schmidhuber J. Framewise phoneme classification with bidirectional LSTM and other neural network architectures. Neural Networks, 2005.
  57. 57.Wang J, Yu L-C, Lai R.K., and Zhang X. Dimensional sentiment analysis using a regional CNN-LSTM model. In Proceedings of the Annual Meeting of the Association for Computational Linguistics (ACL 2016), 2016.
  58. 58.Wang X, Jiang W, Luo Z. Combination of convolutional and recurrent neural network for sentiment analysis of short texts. In Proceedings of the International Conference on Computational Linguistics (COLING 2016), 2016.
  59. 59.Guggilla C, Miller T, Gurevych I. CNN-and LSTM-based claim classification in online user comments. In Proceedings of the International Conference on Computational Linguistics (COLING 2016), 2016.
  60. 60.Huang M, Qian Q, Zhu X. Encoding syntactic knowledge in neural networks for sentiment classification. ACM Transactions on Information Systems, 2017
  61. 61.Akhtar MS, Kumar A, Ghosal D, Ekbal A, and Bhattacharyya P. A multilayer perceptron based ensemble technique for fine-grained financial sentiment analysis. In Proceedings of the Conference on Empirical Methods on Natural Language Processing (EMNLP 2017), 2017.
  62. 62.Guan Z, Chen L, Zhao W, Zheng Y, Tan S, and Cai D. Weakly-supervised deep learning for customer review sentiment classification. In Proceedings of the International Joint Conference on Artificial Intelligence (IJCAI 2016), 2016.
  63. 63.Teng Z, Vo D-T, and Zhang Y. Context-sensitive lexicon features for neural sentiment analysis. In Proceedings of the Conference on Empirical Methods in Natural Language Processing (EMNLP 2016), 2016.
  64. 64.Yu J, Jiang J. Learning sentence embeddings with auxiliary tasks for cross-domain sentiment classification. In Proceedings of the Conference on Empirical Methods in Natural Language Processing (EMNLP 2016), 2016.
  65. 65.Zhao Z, Lu H, Cai D, He X, Zhuang Y. Microblog sentiment classification via recurrent random walk network learning. In Proceedings of the Internal Joint Conference on Artificial Intelligence (IJCAI 2017), 2017.
  66. 66.Mishra A, Dey K, Bhattacharyya P. Learning cognitive features from gaze data for sentiment and sarcasm classification using convolutional neural network. In Proceedings of the Annual Meeting of the Association for Computational Linguistics (ACL 2017), 2017.
  67. 67.Qian Q, Huang M, Lei J, and Zhu X. Linguistically regularized LSTM for sentiment classification. In Proceedings of the Annual Meeting of the Association for Computational Linguistics (ACL 2017), 2017.
  68. 68.Dong L, Wei F, Tan C, Tang D, Zhou M, and Xu K. Adaptive recursive neural network for target-dependent Twitter sentiment classification. In Proceedings of the Annual Meeting of the Association for Computational Linguistics (ACL 2014), 2014.
  69. 69.Vo D-T, Zhang Y. Target-dependent twitter sentiment classification with rich automatic features. In Proceedings of the Internal Joint Conference on Artificial Intelligence (IJCAI 2015), 2015.
  70. 70.Tang D, Qin B, Feng X, and Liu T. Effective LSTMs for target-dependent sentiment classification. In Proceedings of the International Conference on Computational Linguistics (COLING 2016), 2016.
  71. 71.Ruder S, Ghaffari P, Breslin J.G. A hierarchical model of reviews for aspect-based sentiment analysis. In Proceedings of the Conference on Empirical Methods on Natural Language Processing (EMNLP 2016), 2016.
  72. 72.Zhang M, Zhang Y, Vo D-T. Gated neural networks for targeted sentiment analysis. In Proceedings of AAAI Conference on Artificial Intelligence (AAAI 2016), 2016.
  73. 73.Wang Y, Huang M, Zhu X, and Zhao L. Attention-based LSTM for aspect-level sentiment classification. In Proceedings of the Conference on Empirical Methods in Natural Language Processing (EMNLP 2016), 2016.
  74. 74.Yang M, Tu W, Wang J, Xu F, and Chen X. Attention-based LSTM for target-dependent sentiment classification. In Proceedings of AAAI Conference on Artificial Intelligence (AAAI 2017), 2017.
  75. 75.Liu J, Zhang Y. Attention modeling for targeted sentiment. In Proceedings of the Conference of the European Chapter of the Association for Computational Linguistics (EACL 2017), 2017.
  76. 76.Tang D, Qin B, and Liu T. Aspect-level sentiment classification with deep memory network. arXiv preprint arXiv:1605.08900, 2016.
  77. 77.Lei T, Barzilay R, Jaakkola T. Rationalizing neural predictions. In Proceedings of the Conference on Empirical Methods on Natural Language Processing (EMNLP 2016), 2016.
  78. 78.Li C, Guo X, Mei Q. Deep memory networks for attitude Identification. In Proceedings of the ACM International Conference on Web Search and Data Mining (WSDM 2017), 2017.
  79. 79.Ma D, Li S, Zhang X, Wang H. Interactive attention networks for aspect-Level sentiment classification. In Proceedings of the Internal Joint Conference on Artificial Intelligence (IJCAI 2017), 2017.
  80. 80.Chen P, Sun Z, Bing L, and Yang W. Recurrent attention network on memory for aspect sentiment analysis. In Proceedings of the Conference on Empirical Methods on Natural Language Processing (EMNLP 2017), 2017.
  81. 81.Tay Y, Tuan LA, Hui SC. Dyadic memory networks for aspect-based sentiment analysis. In Proceedings of the International Conference on Information and Knowledge Management (CIKM 2017), 2017.
  82. 82.Katiyar A, Cardie C. Investigating LSTMs for joint extraction of opinion entities and relations. In Proceedings of the Annual Meeting of the Association for Computational Linguistics (ACL 2016), 2016.
  83. 83.Wang W, Pan SJ, Dahlmeier D, and Xiao X. Recursive neural conditional random fields for aspect-based sentiment analysis. In Proceedings of the Conference on Empirical Methods in Natural Language Processing (EMNLP 2016), 2016.
  84. 84.Wang W, Pan SJ, Dahlmeier D, and Xiao X. Coupled multi-Layer attentions for co-extraction of aspect and opinion terms. In Proceedings of AAAI Conference on Artificial Intelligence (AAAI 2017), 2017.
  85. 85.Li X, Lam W. Deep multi-task learning for aspect term extraction with memory Interaction. In Proceedings of the Conference on Empirical Methods on Natural Language Processing (EMNLP 2017), 2017.
  86. 86.He R, Lee WS, Ng HT, and Dahlmeier D. An unsupervised neural attention model for aspect extraction. In Proceedings of the Annual Meeting of the Association for Computational Linguistics (ACL 2017), 2017.
  87. 87.Zhang M, Zhang Y, Vo D-T. Neural networks for open domain targeted sentiment. In Proceedings of the Conference on Empirical Methods in Natural Language Processing (EMNLP 2015), 2015.
  88. 88.Zhou X, Wan X, Xiao J. Representation learning for aspect category detection in online reviews. In Proceeding of AAAI Conference on Artificial Intelligence (AAAI 2015), 2015.
  89. 89.Yin Y, Wei F, Dong L, Xu K, Zhang M, and Zhou M. Unsupervised word and dependency path embeddings for aspect term extraction. In Proceedings of the International Joint Conference on Artificial Intelligence (IJCAI 2016), 2016.
  90. 90.Xiong S, Zhang Y, Ji D, and Lou Y. Distance metric learning for aspect phrase grouping. In Proceedings of the International Conference on Computational Linguistics (COLING 2016), 2016.
  91. 91.Poria S, Cambria E, Gelbukh A. Aspect extraction for opinion mining with a deep convolutional neural network. Journal of Knowledge-based Systems. 2016.
  92. 92.Ying D, Yu J, Jiang J. Recurrent neural networks with auxiliary labels for cross-domain opinion target extraction. In Proceedings of AAAI Conference on Artificial Intelligence (AAAI 2017), 2017
  93. 93.Irsoy O, Cardie C. Opinion mining with deep recurrent neural networks. In Proceedings of the Conference on Empirical Methods on Natural Language Processing (EMNLP 2014), 2014.
  94. 94.Yang B, Cardie C. Extracting opinion expressions with semi-markov conditional random fields. In Proceedings of the Conference on Empirical Methods in Natural Language Processing (EMNLP 2012), 2012.
  95. 95.Liu P, Joty S, Meng H. Fine-grained opinion mining with recurrent neural networks and word embeddings. In Proceedings of the Conference on Empirical Methods in Natural Language Processing (EMNLP 2015), 2015.
  96. 96.Irsoy O, Cardie C. Deep recursive neural networks for compositionality in language. In Proceedings of the Annual Conference on Advances in Neural Information Processing Systems (NIPS 2014), 2014.
  97. 97.Zhu X, Guo H, Sobhani P. Neural networks for integrating compositional and non-compositional sentiment in sentiment composition. In Proceedings of the Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (NAACL-HLT 2015), 2015.
  98. 98.Yang B, Cardie C. Joint Inference for fine-grained opinion extraction. In Proceedings of the Annual Meeting of the Association for Computational Linguistics (ACL 2013), 2013.
  99. 99.Deng L, Wiebe J. Recognizing opinion sources based on a new categorization of opinion types. In Proceedings of the International Joint Conference on Artificial Intelligence (IJCAI 2016), 2016.
  100. 100.Chen C, Wang Z, Lei Y, and Li W. Content-based influence modelling for opinion behaviour Prediction. In Proceedings of the International Conference on Computational Linguistics (COLING 2016), 2016.
  101. 101.Rashkin H, Bell E, Choi Y, and Volkova S. Multilingual connotation frames: a case study on social media for targeted sentiment analysis and forecast. In Proceedings of the Annual Meeting of the Association for Computational Linguistics (ACL 2017), 2017.
  102. 102.Mass A. L, Daly R. E, Pham P. T, Huang D, Ng A. Y. and Potts C. Learning word vectors for sentiment analysis. In Proceedings of the Annual Meeting of the Association for Computational Linguistics (ACL 2011), 2011.
  103. 103.Bespalov D, Bai B, Qi Y, and Shokoufandeh A. Sentiment classification based on supervised latent n-gram analysis. In Proceedings of the International Conference on Information and Knowledge Management (CIKM 2011), 2011.
  104. 104.Labutov I, Lipson H. Re-embedding words. In Proceedings of the Annual Meeting of the Association for Computational Linguistics (ACL 2013), 2013.
  105. 105.Tang D, Wei F, Yang N, Zhou M, Liu T, and Qin B. Learning sentiment-specific word embedding for twitter sentiment classification. In Proceedings of the Annual Meeting of the Association for Computational Linguistics (ACL 2014), 2014.
  106. 106.Tang D, Wei F, Qin B, Yang N, Liu T, and Zhoug M. Sentiment embeddings with applications to sentiment analysis. IEEE Transactions on Knowledge and Data Engineering, 2016.
  107. 107.Wang L, Xia R. Sentiment Lexicon construction with representation Learning based on hierarchical sentiment Supervision. In Proceedings of the Conference on Empirical Methods on Natural Language Processing (EMNLP 2017), 2017.
  108. 108.Yu LC, Wang J, Lai KR, and Zhang X. Refining word embeddings for sentiment analysis. In Proceedings of the Conference on Empirical Methods on Natural Language Processing (EMNLP 2017), 2017.
  109. 109.Li J, Jurafsky D. Do multi-sense embeddings improve natural language understanding? In Proceedings of the Conference on Empirical Methods in Natural Language Processing (EMNLP 2015), 2015.
  110. 110.Ren Y, Zhang Y, Zhang, M and Ji D. Improving Twitter sentiment classification using topic-enriched multi-prototype word embeddings. In Proceeding of AAAI Conference on Artificial Intelligence (AAAI 2016), 2016.
  111. 111.Zhou H, Chen L, Shi F, Huang D. Learning bilingual sentiment word embeddings for cross-language sentiment classification. In Proceedings of the Annual Meeting of the Association for Computational Linguistics (ACL 2015), 2015.
  112. 112.Barnes J, Lambert P, Badia T. Exploring distributional representations and machine translation for aspect-based cross-lingual sentiment classification. In Proceedings of the 27th International Conference on Computational Linguistics (COLING 2016), 2016.
  113. 113.Zhang W, Yuan Q, Han J, and Wang J. Collaborative multi-Level embedding learning from reviews for rating prediction. In Proceedings of the International Joint Conference on Artificial Intelligence (IJCAI 2016), 2016.
  114. 114.Sharma R, Somani A, Kumar L, and Bhattacharyya P. Sentiment intensity ranking among adjectives using sentiment bearing word embeddings. In Proceedings of the Conference on Empirical Methods on Natural Language Processing (EMNLP 2017), 2017.
  115. 115.Zhang M, Zhang Y, Fu G. Tweet sarcasm detection using deep neural network. In Proceedings of the International Conference on Computational Linguistics (COLING 2016), 2016.
  116. 116.Joshi A, Tripathi V, Patel K, Bhattacharyya P, and Carman M. Are word embedding-based features useful for sarcasm detection? In Proceedings of the Conference on Empirical Methods on Natural Language Processing (EMNLP 2016), 2016.
  117. 117.Poria S, Cambria E, Hazarika D, and Vij P. A deeper look into sarcastic tweets using deep convolutional neural networks. In Proceedings of the International Conference on Computational Linguistics (COLING 2016), 2016.
  118. 118.Peled L, Reichart R. Sarcasm SIGN: Interpreting sarcasm with sentiment based monolingual machine translation. In Proceedings of the Annual Meeting of the Association for Computational Linguistics (ACL 2017), 2017.
  119. 119.Ghosh A, Veale T. Magnets for sarcasm: making sarcasm detection timely, contextual and very personal. In Proceedings of the Conference on Empirical Methods on Natural Language Processing (EMNLP 2017), 2017.
  120. 120.Van Hee C, Lefever E, Hoste V. Monday mornings are my fave:)# not exploring the automatic recognition of irony in english tweets. In Proceedings of the International Conference on Computational Linguistics (COLING 2016), 2016.
  121. 121.Chen WF, Lin FY, Ku LW. WordForce: visualizing controversial words in debates. In Proceedings of the Conference on Empirical Methods in Natural Language Processing (EMNLP 2016), 2016.
  122. 122.Wang Z, Zhang Y, Lee S, Li S, and Zhou G. A bilingual attention network for code-switched emotion prediction. In Proceedings of the International Conference on Computational Linguistics (COLING 2016), 2016.
  123. 123.Zhou H, Huang M, Zhang T, Zhu X and Liu B. Emotional chatting machine: emotional conversation generation with internal and external memory. arXiv preprint. arXiv:1704.01074, 2017.
  124. 124.Abdul-Mageed M, Ungar L. EmoNet: fine-grained emotion detection with gated recurrent neural networks. In Proceedings of the Annual Meeting of the Association for Computational Linguistics (ACL 2017), 2017.
  125. 125.Felbo B, Mislove A, Søgaard A, Rahwan I, and Lehmann S. Using millions of emoji occurrences to learn any-domain representations for detecting sentiment, emotion and sarcasm. In Proceedings of the Conference on Empirical Methods on Natural Language Processing (EMNLP 2017), 2017.
  126. 126.Gui L, Hu J, He Y, Xu R, Lu Q, and Du J. A question answering approach to emotion cause extraction. In Proceedings of the Conference on Empirical Methods on Natural Language Processing (EMNLP 2017), 2017.
  127. 127.Poria S, Cambria E, Gelbukh A. Deep convolutional neural text features and multiple kernel learning for utterance-level multimodal sentiment analysis. In Proceedings of the Conference on Empirical Methods on Natural Language Processing (EMNLP 2015), 2015.
  128. 128.Bertero D, Siddique FB, Wu CS, Wan Y, Chan R.H, and Fung P. Real-time speech emotion and sentiment recognition for interactive dialogue systems. In Proceedings of the Conference on Empirical Methods in Natural Language Processing (EMNLP 2016), 2016.
  129. 129.Fung P, Dey A, Siddique FB, Lin R, Yang Y, Bertero D, Wan Y, Chan RH, and Wu CS. Zara: a virtual interactive dialogue system incorporating emotion, sentiment and personality recognition. In Proceedings of the International Conference on Computational Linguistics (COLING 2016), 2016.
  130. 130.Wang J, Fu J, Xu Y, and Mei T. Beyond object recognition: visual sentiment analysis with deep coupled adjective and noun neural networks. In Proceedings of the Internal Joint Conference on Artificial Intelligence (IJCAI 2016), 2016.
  131. 131.Yang J, Sun M, Sun X. Learning visual sentiment distributions via augmented conditional probability neural network. In Proceedings of AAAI Conference on Artificial Intelligence (AAAI 2017), 2017.
  132. 132.Zhu X, Li L, Zhang W, Rao T, Xu M, Huang Q, and Xu D. Dependency exploitation: a unified CNN-RNN approach for visual emotion recognition. In Proceedings of the Internal Joint Conference on Artificial Intelligence (IJCAI 2017), 2017.
  133. 133.You Q, Jin H, Luo J. Visual sentiment analysis by attending on local image regions. In Proceedings of AAAI Conference on Artificial Intelligence (AAAI 2017), 2017.
  134. 134.Poria S, Cambria E, Hazarika D, Majumder N, Zadeh A, and Morency LP. Context-dependent sentiment analysis in user-generated videos. In Proceedings of the Annual Meeting of the Association for Computational Linguistics (ACL 2017), 2017.
  135. 135.Tripathi S, Acharya S, Sharma RD, Mittal S, and Bhattacharya S. Using deep and convolutional neural networks for accurate emotion classification on DEAP dataset. In Proceedings of AAAI Conference on Artificial Intelligence (AAAI 2017), 2017.
  136. 136.Zadeh A, Chen M, Poria S, Cambria E, and Morency LP. Tensor fusion network for multimodal sentiment analysis. In Proceedings of the Conference on Empirical Methods on Natural Language Processing (EMNLP 2017), 2017.
  137. 137.Long Y, Qin L, Xiang R, Li M, and Huang CR. A cognition based attention model for sentiment analysis. In Proceedings of the Conference on Empirical Methods on Natural Language Processing (EMNLP 2017), 2017.
  138. 138.Wang H, Meghawat A, Morency LP, and Xing E.X. Select-additive learning: improving generalization in multimodal sentiment analysis. In Proceedings of the International Conference on Multimedia and Expo (ICME 2017), 2017.
  139. 139.Akhtar MS, Kumar A, Ekbal A, and Bhattacharyya P. A hybrid deep learning architecture for sentiment analysis. In Proceedings of the International Conference on Computational Linguistics (COLING 2016), 2016.
  140. 140.Dahou A, Xiong S, Zhou J, Haddoud MH, and Duan P. Word embeddings and convolutional neural network for Arabic sentiment classification. In Proceedings of the International Conference on Computational Linguistics (COLING 2016), 2016.
  141. 141.Singhal P, Bhattacharyya P. Borrow a little from your rich cousin: using embeddings and polarities of english words for multilingual sentiment classification. In Proceedings of the International Conference on Computational Linguistics (COLING 2016), 2016.
  142. 142.Joshi A, Prabhu A, Shrivastava M, and Varma V. Towards sub-word level compositions for sentiment analysis of Hindi-English code mixed text. In Proceedings of the International Conference on Computational Linguistics (COLING 2016), 2016.
  143. 143.Gui L, Xu R, He Y, Lu Q, and Wei Z. Intersubjectivity and sentiment: from language to knowledge. In Proceedings of the International Joint Conference on Artificial Intelligence (IJCAI 2016), 2016.
  144. 144.Wang Y, Zhang Y, Liu B. Sentiment lexicon expansion based on neural PU Learning, double dictionary lookup, and polarity association. In Proceedings of the Conference on Empirical Methods on Natural Language Processing (EMNLP 2017), 2017.
  145. 145.Rekabsaz N, Lupu M, Baklanov A, Hanbury A, Dür A, and Anderson L. Volatility prediction using financial disclosures sentiments with word embedding-based IR models. In Proceedings of the Annual Meeting of the Association for Computational Linguistics (ACL 2017), 2017.
  146. 146.Wang Z, Zhang Y. Opinion recommendation using a neural model. In Proceedings of the Conference on Empirical Methods on Natural Language Processing (EMNLP 2017), 2017.
  147. 147.Augenstein I, Rocktäschel T, Vlachos A, Bontcheva K. Stance detection with bidirectional conditional encoding. In Proceedings of the Conference on Empirical Methods in Natural Language Processing (EMNLP 2016), 2016.
  148. 148.Du J, Xu R, He Y, Gui L. Stance classification with target-specific neural attention networks. In Proceedings of the Internal Joint Conference on Artificial Intelligence (IJCAI 2017), 2017.

Citation

MLA
Zhang, L., et al. “Deep Learning for Sentiment Analysis : A Survey”. arXiv, 2018, http://arxiv.org/abs/1801.07883v2.
APA
Zhang, L., Wang, S., & Liu, B. (2018). Deep Learning for Sentiment Analysis : A Survey. arXiv. http://arxiv.org/abs/1801.07883v2
Chicago
Zhang, L., S. Wang, and B. Liu. 2018. “Deep Learning for Sentiment Analysis : A Survey”. arXiv. http://arxiv.org/abs/1801.07883v2.
Harvard
Zhang, L., Wang, S. and Liu, B. (2018) “Deep Learning for Sentiment Analysis : A Survey”, arXiv [Preprint]. Available at: http://arxiv.org/abs/1801.07883v2.
Vancouver
1. Zhang L, Wang S, Liu B (2018) Deep Learning for Sentiment Analysis : A Survey. arXiv

BibTeX

@article{zhang2018deep,
  title = {Deep Learning for Sentiment Analysis : A Survey},
  author = {Zhang, Lei and Wang, Shuai and Liu, Bing},
  year = {2018},
  journal = {arXiv},
  url = {http://arxiv.org/abs/1801.07883v2},
  eprint = {1801.07883}
}
Metadata:arXiv

Access the Paper

This paper is available from its original source. Click below to access the PDF.

Open PDF