Thumbs Up or Thumbs Down? Semantic Orientation Applied to Unsupervised Classification of Reviews

Abstract

This paper presents a simple unsupervised learning algorithm for classifying reviews as recommended (thumbs up) or not recommended (thumbs down). The classification of a review is predicted by the average semantic orientation of the phrases in the review that contain adjectives or adverbs. A phrase has a positive semantic orientation when it has good associations (e.g., "subtle nuances") and a negative semantic orientation when it has bad associations (e.g., "very cavalier"). In this paper, the semantic orientation of a phrase is calculated as the mutual information between the given phrase and the word "excellent" minus the mutual information between the given phrase and the word "poor". A review is classified as recommended if the average semantic orientation of its phrases is positive. The algorithm achieves an average accuracy of 74% when evaluated on 410 reviews from Epinions, sampled from four different domains (reviews of automobiles, banks, movies, and travel destinations). The accuracy ranges from 84% for automobile reviews to 66% for movie reviews.

8 pages

Cited by in corpus (25)

• Wikipedia-based Semantic Interpretation for Natural Language Processing
• Empath: Understanding Topic Signals in Large-Scale Text
• Sentiment Analysis of Twitter Data: A Survey of Techniques
• A Survey on Aspect-Based Sentiment Classification
• A Machine Learning Approach For Opinion Holder Extraction In Arabic Language
• Sentiment Analysis For Modern Standard Arabic And Colloquial
• TensiStrength: Stress and relaxation magnitude detection for social media texts
• MoralStrength: Exploiting a Moral Lexicon and Embedding Similarity for Moral Foundations Prediction
• Multi-modal Sarcasm Detection and Humor Classification in Code-mixed Conversations
• A Case Study of Spanish Text Transformations for Twitter Sentiment Analysis
• A Survey of Data Mining Techniques for Social Media Analysis
• Understanding Environmental Posts: Sentiment and Emotion Analysis of Social Media Data
• Deep Learning-based Sentiment Classification: A Comparative Survey
• Jointly identifying opinion mining elements and fuzzy measurement of opinion intensity to analyze product features
• Automatic Aggregation by Joint Modeling of Aspects and Values
• The impact of political party/candidate on the election results from a sentiment analysis perspective using #AnambraDecides2017 tweets
• Which Clustering Do You Want? Inducing Your Ideal Clustering with Minimal Feedback
• Lancaster A at SemEval-2017 Task 5: Evaluation metrics matter: predicting sentiment from financial news headlines
• Combination of Domain Knowledge and Deep Learning for Sentiment Analysis
• Idioms-Proverbs Lexicon for Modern Standard Arabic and Colloquial Sentiment Analysis
• Integrating sentiment and social structure to determine preference alignments: The Irish Marriage Referendum
• Mining Software Quality from Software Reviews: Research Trends and Open Issues
• Semantic Properties of Customer Sentiment in Tweets
• Combination of Domain Knowledge and Deep Learning for Sentiment Analysis of Short and Informal Messages on Social Media
• Sentiment Analysis with Contextual Embeddings and Self-Attention