The Use of AI to Measure Gross National Happiness

This research attempts to identify an alternative approach to measuring Gross National Happiness (GNH).It uses artificial intelligence (AI), incorporating natural language processing (NLP) and sentiment analysis to measure GNH.It uses four widely used AI approaches to estimate GNH scores from online content such as social media posts and news websites to provide a cost-practical alternative approach to the traditional approach of GNH measurement.Traditional GNH measurement uses expensive and time-consuming surveys that take place infrequently, once every few years.This research addresses some of the limitations of conventional surveys.We use existing 'off the shelf ' NLP models to encode social media-based text data and then conduct sentiment analysis, testing several AI classification approaches to estimate GNH scores.The four classification approaches tested to estimate the sentiment scores are the Multi-layer Perceptron (MLP) classifier, decision tree, linear model, and K-nearest neighbors (K-NN).The results show the MLP approach to achieve the highest accuracy on the test data set.The research concludes that this approach can estimate a country's GNH scores more cost-effectively and timely than the traditional approach.Applying AI on social media and news websites provides a real-time understanding of society's progress, concerns, and challenges.This can help cost-effectively influence effective policy-making.A pilot study was conducted to measure the GNH of the Royal Kingdom of Bhutan.