Measuring a Texts Fairness Dimensions Using Machine Learning Based on Social Psychological Factors

Abstract

Fairness is a principal social value that can be observed in civilisations around the world. A manifestation of this is in social agreements, often described in texts, such as contracts. Yet, despite the prevalence of such, a fairness metric for texts describing a social act remains wanting. To address this, we take a step back to consider the problem based on first principals. Instead of using rules or templates, we utilise social psychology literature to determine the principal factors that humans use when making a fairness assessment. We then attempt to digitise these using word embeddings into a multi-dimensioned sentence level fairness perceptions vector to serve as an approximation for these fairness perceptions. The method leverages a pro-social bias within word embeddings, for which we obtain an F1= 81.0. A second approach, using PCA and ML based on the said fairness approximation vector produces an F1 score of 86.2. We detail improvements that can be made in the methodology to incorporate the projection of sentence embedding on to a subspace representation of fairness.

arXiv comment38 pages, 9 figures Change in Author details. Added acknowledgment section

References in corpus (11)

• Universal Sentence Encoder
• Word Embeddings Quantify 100 Years of Gender and Ethnic Stereotypes
• The Measure and Mismeasure of Fairness
• Efficient Natural Language Response Suggestion for Smart Reply
• A Survey of the State of Explainable AI for Natural Language Processing
• Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks
• MoralStrength: Exploiting a Moral Lexicon and Embedding Similarity for Moral Foundations Prediction
• Efficient Vector Representation for Documents through Corruption
• Relating Word Embedding Gender Biases to Gender Gaps: A Cross-Cultural Analysis
• BERT has a Moral Compass: Improvements of ethical and moral values of machines
• Unsupervised Document Embedding With CNNs