Signal 1: What Artificial Intelligence Reveals About Urban Change
Although monitoring urban changes has been essential factor for urban planners and policymakers to find out where city problem is located to build up resolution, it has been challenging because of analysis methodology. To monitor urban change in traditional way, quantitative data such as total building data change which represents new construction is used to look for the site where city changes occur. However, this process is very labor-intensive task and might include error unless analyzing unit is building lot. For example, government agency have to collect data in advance and data cleaning process should be completed. Thus, total change might not include information about reality depending on analysis unit.
Recently, as we are living in big data era, new methodology such as machine learning has been suggested that might help for solving city problems. The biggest benefit of machine learning is that let us be able to analyze qualitative data like natural language, sound and imagery. In the article, author introduce the research that try to find the place that changes occur(development or demolition) via machine learning with Google Street View Image data. The trained model recognize different figure in the image of location between past and present so that researcher can find the place that urban change occur. The study shows that such new skills can play an important role in analyzing and monitoring urban physical environments.
