Data Driven Urban Data Mining
Event Type
Student Poster
TimeTuesday, July 306:30pm - 8:30pm
LocationCrystal Foyer and Crystal B
DescriptionUrban planning seeks to use big data to build smart cities. Urban data in NewMexico was crawled from creditable sources on-line and by requesting. Twocase studies were conducted in this research for the purpose of urban plan-ning and address two questions in our research. For case study 1, all the datawas collected fromNew Mexico Environmental Public Health Tracking, such as,Asthma, COPD, and Ozone, which are the major factors focused in this researchand case study II requested data from theNew Mexico Department of Trans-portation. Case study I, processed the data using programming language ofpython by implementing imported library,pandas. The data was inputted intoscatter plots to find correlation between Ozone and Asthma, Ozone and COPD.However, the plots did not show much promising correlation, so Pearson corre-lation and K-means clusters are the analysis tools used to test whether there is acorrelation between these factors. Case study II used python imported librariespandas, sklearn, and seaborn to visualize the data spatially and by time. Trendsfound through visualization was used as a basis for preprocessing of data andthen running the apriori algorithm to determine highest frequent combinationsof factors. Future study will involve crawling of more factors to include in spa-tial analysis to help better understand the trends and peaks seen in the data.Analysis of the factors will provide invaluable information to the Departmentof Transportation and first responders in the area.