Can websites reveal the extent and degree to which a business’s values reflect national policy? A text embeddings approach

Hogan, AlexanderMoran, StephanieHogan, KevinBarker, BethWoodall, Richard2024-09-202024-09-202024-07-169788413962016https://riunet.upv.es/handle/10251/208403[EN] This paper demonstrates a novel application of WebAI, applying word vectorisation to website texts to measure the prevalence and proliferation of values associated with the Well-being of Future Generations (Wales) Act 2015 amongst Welsh businesses. This was achieved by training a model to recognise topics of particular importance in texts describing the WFGA, and then using this model to analyse the text hosted on Welsh business’s websites. The companies studied comprised a structured national sample of the Welsh economy. Our findings indicate that the method detects meaningful differences between regions, business sector and participants in a government innovation support program.9Reconocimiento - No comercial - Compartir igual (by-nc-sa)EconometricsSociometricsPolicy evaluationBig data analyticsText embeddingsVectorisationCan websites reveal the extent and degree to which a business’s values reflect national policy? A text embeddings approachCapítulo de libro10.4995/CARMA2024.2024.17801Abierto