Text Mining in Smart Cities to Identify Urban Events and Public Service Problems

摘要

Cities will be constituted in large agglomerations and will be the axis of social, economic, cultural, and artistic human activity. According to the latest United Nations reports, in 2050, the cities will concentrate on 68% of the world's population; this means that progressively the world will no longer be rural and will become urban. Local governments can benefit of the ICTs in order to collect data and make smarter decision and policy making for such large cities. The Smart Cities will be related to the use of social networks and social mining to be able to analyze the opinions of citizens and their reactions to the governors or governmental entities. Currently, social networks such as Twitter have provided the opportunity to perform analysis on the user interactions and sentiments. We intend to apply a tweets text mining process to identify public service problems and urban events related, for instance, traffic and security. A case study in Ecuador capital Quito and its metropolitan area. The present research allows identifying, on the one hand, the inconveniences/problems related to public services and urban events such as water, electricity, mobility, public transport, and security at the level of different areas. A temporal and spatial identification of such problems is also carried out.