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

2005/06/17 by Bo Pang, Lillian Lee, Pang, Bo +1 · 58 citations
Computer Science · #Advanced Text Analysis Techniques #Sentiment Analysis and Opinion Mining #Text and Document Classification Technologies #cs.CL #cs.LG

paper · pdf · doi:10.48550/arxiv.cs/0506075

To appear, Proceedings of ACL 2005

arxiv created 2005/06/17 · arxiv updated 2009/12/01

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.

Cited by

• A Coulomb Particle Model for Learning Kernel Attention in Transformers
• Retrieval-augmented Prompt Learning for Pre-trained Foundation Models
• MAGIC: Achieving Superior Model Merging via Magnitude Calibration
• Convolutional Lie Operator for Sentence Classification
• Let the Barbarians In: How AI Can Accelerate Systems Performance Research
• PIAST: Rapid Prompting with In-context Augmentation for Scarce Training data
• Technical Report on Text Dataset Distillation
• Masks Can Be Distracting: On Context Comprehension in Diffusion Language Models
• Geometry of Decision Making in Language Models
• GPS: General Per-Sample Prompter
• Curriculum CycleGAN for Textual Sentiment Domain Adaptation with Multiple Sources
• Tighter Truncated Rectangular Prism Approximation for RNN Robustness Verification
• FLEX: Unifying Evaluation for Few-Shot NLP
• Llama-Embed-Nemotron-8B: A Universal Text Embedding Model for Multilingual and Cross-Lingual Tasks
• Any-gram Kernels for Sentence Classification: A Sentiment Analysis Case\n Study
• FedRW: Efficient Privacy-Preserving Data Reweighting for Enhancing Federated Learning of Language Models
• Supervised Fine Tuning for Word Embedding with Integrated Knowledge
• Calibration Across Layers: Understanding Calibration Evolution in LLMs
• Linguistically Regularized LSTMs for Sentiment Classification
• Implicit Regularization of Stochastic Gradient Descent in Natural Language Processing: Observations and Implications
• Disentangled Representations for Manipulation of Sentiment in Text
• Hierarchical Bi-Directional Self-Attention Networks for Paper Review Rating Recommendation
• Convolutional Neural Networks for Sentence Classification
• Learning Generic Sentence Representations Using Convolutional Neural Networks
• An efficient framework for learning sentence representations
• Self-Balanced Dropout
• SentEval: An Evaluation Toolkit for Universal Sentence Representations
• Evaluation of sentence embeddings in downstream and linguistic probing\n tasks
• SimCSE: Simple Contrastive Learning of Sentence Embeddings
• Multitask Prompted Training Enables Zero-Shot Task Generalization
• Adaptive Blockwise Search: Inference-Time Alignment for Large Language Models
• Rethinking Skip-thought: A Neighborhood based Approach
• Investigating Capsule Networks with Dynamic Routing for Text Classification
• Barbarians at the Gate: How AI is Upending Systems Research
• Does Using Counterfactual Help LLMs Explain Textual Importance in Classification?
• Submodular Context Partitioning and Compression for In-Context Learning
• Localizing Task Recognition and Task Learning in In-Context Learning via Attention Head Analysis
• GuardNet: Graph-Attention Filtering for Jailbreak Defense in Large Language Models
• Train Once, Reuse Everywhere: Generalizable Implicit In-Context Learning by Routing Attention
• Binary Autoencoder for Mechanistic Interpretability of Large Language Models
• Improving the Accuracy of Pre-trained Word Embeddings for Sentiment Analysis
• Is BERT Really Robust? A Strong Baseline for Natural Language Attack on Text Classification and Entailment
• BASFuzz: Towards Robustness Evaluation of LLM-based NLP Software via Automated Fuzz Testing
• Learning Semantically Coherent and Reusable Kernels in Convolution Neural Nets for Sentence Classification
• RanAT4BIE: Random Adversarial Training for Biomedical Information Extraction
• Graph Convolutional Networks for Text Classification
• Dependency-based Convolutional Neural Networks for Sentence Embedding
• Research Challenges in Relational Database Management Systems for LLM Queries
• Data Representation and Compression Using Linear-Programming\n Approximations
• Universal Sentence Encoder
• Cutting Recursive Autoencoder Trees
• Memory Decoder: A Pretrained, Plug-and-Play Memory for Large Language Models
• Character-level Convolutional Network for Text Classification Applied to Chinese Corpus
• Evaluating Selective Encryption Against Gradient Inversion Attacks
• Cross-lingual Opinions and Emotions Mining in Comparable Documents
• MLP Memory: A Retriever-Pretrained Memory for Large Language Models
• Adversarial Defence without Adversarial Defence: Enhancing Language Model Robustness via Instance-level Principal Component Removal
• Uncovering Gradient Inversion Risks in Practical Language Model Training

Related

• KNNs of Semantic Encodings for Rating Prediction
• A New Approach To Text Rating Classification Using Sentiment Analysis
• Determining sentiment in citation text and analyzing its impact on the proposed ranking index
• Thumbs up? Sentiment Classification using Machine Learning Techniques
• Correlation-Based Method for Sentiment Classification