I am studying the emergence of new social norms in food consumption. New social norms seems to be emergent in industrialized countries (vegetarianism, vegan). How would you measure in text data (newspaper, scientific papers, grey literature, etc) such emergence?
I would interpret your text data as opinion texts and measure emergence by performing opinion mining. IMHO, emergence of norms highly correlates with intensity of opinions and you can easily adopt existing approaches to opinion mining in your texts.
Schematically, it would work as follows. Given that you find many emotion words in a norm text, this can be a cue of high emergence of a norm. Of course, it is a simple description of how you can measure emergence and you shouldn't do it using word spotting if you have long texts -- instead of word spotting for measuring emergence you should consider using a statistical approach, for instance, that uses SVM or NaiveBayes (statistical approaches are more reliable if they are applied to long data).
The result that you get from a mathematical algorithm might be not very self-explanatory since you get a confusion matrix with many numbers. You might be interested in result interpretation (s. the link).
Thesis Opinion Mining and Lexical Affect Sensing
Distributed Emergent Cultural Cognition {DECC} (Sharifian, 2008)
https://pdfs.semanticscholar.org/050d/87fc38ed31eb2afb964039c5e99938f36904.pdf
- Badges
- Science topic