Public opinion is nowadays a valuable data source for many sectors. We analyzed the #transportation sector using messages extracted from #Twitter. Contrasting with the traditional surveying methods that are high-cost and inefficient used in the transportation sector, #socialmedia are popular sources of crowdsensing. We used BERT #embeddings, an unsupervised pre-trained model released in 2018, to classify travel-related terms using tweets collected from three distinct cities: New York, London, and Melbourne. This research work was presented today at the International Smart City Conference (#isc2).
#optiMOVES, #INESCTEC, #urbanmobility, #transport, #transport, #smartcities
