Seeing stars: Exploiting class relationships for sentiment categorization with respect to rating scales

Abstract

We address the rating-inference problem, wherein rather than simply decide whether a review is "thumbs up" or "thumbs down", as in previous sentiment analysis work, one must determine an author's evaluation with respect to a multi-point scale (e.g., one to five "stars"). This task represents an interesting twist on standard multi-class text categorization because there are several different degrees of similarity between class labels; for example, "three stars" is intuitively closer to "four stars" than to "one star". We first evaluate human performance at the task. Then, we apply a meta-algorithm, based on a metric labeling formulation of the problem, that alters a given n-ary classifier's output in an explicit attempt to ensure that similar items receive similar labels. We show that the meta-algorithm can provide significant improvements over both multi-class and regression versions of SVMs when we employ a novel similarity measure appropriate to the problem.

To appear, Proceedings of ACL 2005

References in corpus (1)

• Attention-Sensitive Alerting

Cited by in corpus (120)

• Universal Sentence Encoder
• Convolutional Neural Networks for Sentence Classification
• Unsupervised Learning of Sentence Embeddings using Compositional n-Gram Features
• Multitask Prompted Training Enables Zero-Shot Task Generalization
• An efficient framework for learning sentence representations
• Learning to Generate Reviews and Discovering Sentiment
• SentEval: An Evaluation Toolkit for Universal Sentence Representations
• TextBugger: Generating Adversarial Text Against Real-world Applications
• Modeling public mood and emotion: Twitter sentiment and socio-economic phenomena
• Adversarial Attacks on Deep Learning Models in Natural Language Processing: A Survey
• Adversarial Training Methods for Semi-Supervised Text Classification
• Get out the vote: Determining support or opposition from Congressional floor-debate transcripts
• SentiWords: Deriving a High Precision and High Coverage Lexicon for Sentiment Analysis
• Evaluation of sentence embeddings in downstream and linguistic probing tasks
• A Survey on Text Classification: From Shallow to Deep Learning
• Investigating Capsule Networks with Dynamic Routing for Text Classification
• Is BERT Really Robust? A Strong Baseline for Natural Language Attack on Text Classification and Entailment
• Linguistically Regularized LSTMs for Sentiment Classification
• Learning Supervised Topic Models for Classification and Regression from Crowds
• Adversarial Multi-task Learning for Text Classification
• Multi-modal Sarcasm Detection and Humor Classification in Code-mixed Conversations
• Dependency-based Convolutional Neural Networks for Sentence Embedding
• Text Classification using Capsules
• A Survey on Sentiment and Emotion Analysis for Computational Literary Studies
• Can Adversarial Weight Perturbations Inject Neural Backdoors?
• Improving the Accuracy of Pre-trained Word Embeddings for Sentiment Analysis
• Graph Convolutional Networks for Text Classification
• Universal, Unsupervised (Rule-Based), Uncovered Sentiment Analysis
• Prompt as Triggers for Backdoor Attack: Examining the Vulnerability in Language Models
• Stein Variational Gradient Descent With Matrix-Valued Kernels
• Improving Review Representations with User Attention and Product Attention for Sentiment Classification
• The supervised hierarchical Dirichlet process
• Word Embedding Perturbation for Sentence Classification
• A Multi-task Ensemble Framework for Emotion, Sentiment and Intensity Prediction
• Machine Learning Models that Remember Too Much
• Visualizing and Understanding Curriculum Learning for Long Short-Term Memory Networks
• Pair-Level Supervised Contrastive Learning for Natural Language Inference
• "That Is a Suspicious Reaction!": Interpreting Logits Variation to Detect NLP Adversarial Attacks
• Understanding Convolutional Neural Networks for Text Classification
• Lexicon Integrated CNN Models with Attention for Sentiment Analysis
• Statistical Inferences for Polarity Identification in Natural Language
• AC-BLSTM: Asymmetric Convolutional Bidirectional LSTM Networks for Text Classification
• KATE: K-Competitive Autoencoder for Text
• Character-level Convolutional Network for Text Classification Applied to Chinese Corpus
• Debiased Contrastive Learning
• Learning Generic Sentence Representations Using Convolutional Neural Networks
• Exploring Shared Structures and Hierarchies for Multiple NLP Tasks
• A Data-driven Neural Network Architecture for Sentiment Analysis
• Bayesian Sparsification of Recurrent Neural Networks
• Conditional BERT Contextual Augmentation
• Deep Active Learning for Text Classification with Diverse Interpretations
• Generating Natural Language Adversarial Examples on a Large Scale with Generative Models
• Active Discriminative Text Representation Learning
• Recursive Graphical Neural Networks for Text Classification
• Distributed Representations for Compositional Semantics
• Integrating sentiment and social structure to determine preference alignments: The Irish Marriage Referendum
• Dynamic Compositional Neural Networks over Tree Structure
• On the Evaluation of the Plausibility and Faithfulness of Sentiment Analysis Explanations
• Disentangled Representations for Manipulation of Sentiment in Text
• ParVecMF: A Paragraph Vector-based Matrix Factorization Recommender System
• Implicit Regularization of Stochastic Gradient Descent in Natural Language Processing: Observations and Implications
• Short Text Classification via Knowledge powered Attention with Similarity Matrix based CNN
• Dependency Sensitive Convolutional Neural Networks for Modeling Sentences and Documents
• Improving Visually Grounded Sentence Representations with Self-Attention
• An Attention-Gated Convolutional Neural Network for Sentence Classification
• Densely Connected Bidirectional LSTM with Applications to Sentence Classification
• Multi-view Sentence Representation Learning
• FLEX: Unifying Evaluation for Few-Shot NLP
• Learning Semantically Coherent and Reusable Kernels in Convolution Neural Nets for Sentence Classification
• Performance Investigation of Feature Selection Methods
• Detecting state of aggression in sentences using CNN
• Leveraging Sparse and Dense Feature Combinations for Sentiment Classification
• SeVeN: Augmenting Word Embeddings with Unsupervised Relation Vectors
• When Do Discourse Markers Affect Computational Sentence Understanding?
• Prediction-Constrained Training for Semi-Supervised Mixture and Topic Models
• Supervised Fine Tuning for Word Embedding with Integrated Knowledge
• Preserving Semantics in Textual Adversarial Attacks
• Textual Membership Queries
• Predicting the Sentiment Polarity and Rating of Yelp Reviews
• Improving Sentence Representations with Consensus Maximisation
• Cutting Recursive Autoencoder Trees
• Evaluation of Unsupervised Compositional Representations
• Sentiment Analysis on Speaker Specific Speech Data
• Contextual Recurrent Units for Cloze-style Reading Comprehension
• Rethinking Skip-thought: A Neighborhood based Approach
• Automatic Identification of Sarcasm Target: An Introductory Approach
• Empirical Evaluation of RNN Architectures on Sentence Classification Task
• Analyzing Multi-Task Learning for Abstractive Text Summarization
• Curriculum CycleGAN for Textual Sentiment Domain Adaptation with Multiple Sources
• Sentiment Analysis by Joint Learning of Word Embeddings and Classifier
• Multiple Instance Learning Networks for Fine-Grained Sentiment Analysis
• Towards Interpreting Recurrent Neural Networks through Probabilistic Abstraction
• Estimating the Rating of Reviewers Based on the Text
• Translations as Additional Contexts for Sentence Classification
• Detecting Low Rating Android Apps Before They Have Reached the Market
• Any-gram Kernels for Sentence Classification: A Sentiment Analysis Case Study
• R&R: Metric-guided Adversarial Sentence Generation
• Data Representation and Compression Using Linear-Programming Approximations
• A Novel Framework for Recurrent Neural Networks with Enhancing Information Processing and Transmission between Units
• Learning Universal Sentence Representations with Mean-Max Attention Autoencoder
• Applying SoftTriple Loss for Supervised Language Model Fine Tuning
• Set Aggregation Network as a Trainable Pooling Layer
• Dynamic Compositionality in Recursive Neural Networks with Structure-aware Tag Representations
• Identity-sensitive Word Embedding through Heterogeneous Networks
• Mixture of Expert/Imitator Networks: Scalable Semi-supervised Learning Framework
• Aggregated Learning: A Deep Learning Framework Based on Information-Bottleneck Vector Quantization
• Deep Natural Language Processing for LinkedIn Search
• Domain-adaptation of spherical embeddings
• Hierarchical Bi-Directional Self-Attention Networks for Paper Review Rating Recommendation
• Learning to Compose over Tree Structures via POS Tags
• Deep Active Learning by Model Interpretability
• Metric Learning for Dynamic Text Classification
• Understanding Negations in Information Processing: Learning from Replicating Human Behavior
• Learning with Clustering Structure
• A Novel Large-scale Ordinal Regression Model
• GWU NLP Lab at SemEval-2019 Task 3: EmoContext: Effective Contextual Information in Models for Emotion Detection in Sentence-level in a Multigenre Corpus
• Exploiting Domain Knowledge via Grouped Weight Sharing with Application to Text Categorization
• Task-specific Word Identification from Short Texts Using a Convolutional Neural Network
• Revisiting the Importance of Encoding Logic Rules in Sentiment Classification
• Self-Balanced Dropout