Researchers examined the language used in tweets from 1,300 US counties and found it to be predictive of the subjective well-being of people living in those counties. Using LDA to derive topics, sets of cooccurring words, the researchers found that these topics improved the accuracy of predicting life satisfaction beyond standard demographic and socio-economic controls. The LDA topics provide a greater behavioural and conceptual resolution into life satisfaction than the broad socio-economic and demographic variables, allowing for more specific insights into the factors that influence well-being. For example, words relating to outdoor activities, spiritual meaning, exercise, and good jobs were found to correlate with increased life satisfaction, while words signifying disengagement such as 'bored' and 'tired' showed a negative association. These findings suggest that social media language analysis could offer a cost-effective, timely, and broadly applicable way to measure the well-being of populations.
Read the research: Characterizing Geographic Variation in Well-Being Using Tweets
