Methodology and mapping

Data - The data collected was essential to creating the mappings. When defining the scope of the research, I had defined what the borders of the map were. Further, I continued to define the scale needed for the maps to be appropriate. Further borders were drawn in how the city had organized its neighborhoods into ZIP codes. These allowed for a sizeable amount of specific points within the map to be generalized. The center point of the ZIP code was to be the coordinate used for the distance measures. Geolocating the permanent HIV-providers was the second important set of data. These were determined points on the map since they existed within buildings in the city with detailed coordinates. I continued by finding the location of the metro stops and the paths in which they connect with different metro lines. Lastly, I had to compile datasets with coordinates or ZIP codes of the socioeconomic background within Mexico City in order to use it for comparison. 

Methodology - In compiling these datasets together, and using ARCGIS, the time needed to traverse certain distances by using the metro is calculated. By using the sequential choropleth method of visualization, each ZIP code area is assigned a tone between white and black—determining the time (in minutes) associated with traveling to the HIV service providers.  This allows for a gradient to develop, with white being the closest to the providers, and black with furthest away. This gradient allows for a margin of error that compensates for the defined number of points in the map provided by the ZIP codes. Most importantly, the methodology is focused on determining distance by time needed for a round trip to the HIV service providers. In the scale of Mexico City, this lead to times that ranged up to five hours, without including the waiting times of the transportation or of the HIV service providers and appointments. 

Mapping - In applying the methodology and synthesizing the dataset, I began to finalize my first visual representation of the accessibility to HIV service providers in Mexico City by public transportation. While doing this, I explored different ways in which other research has depicted similar issues. The gradient determined by the choropleth method seemed most feasible, though it brought up theoretical questions of neighborhood borders and in-between city spaces. Furthermore, I dealt with issues of representation of small scale, since the scale of the city did not compare to that of singular points like metro stops and buildings. Though, in compiling all the information together, the map began to clearly show the discrepancy between different districts within Mexico City. There seemed to be clusters of HIV service providers, which did not particularly correlate with the density of population that lived in the area. I wanted the first map to have a clarity in language, no overcomplication in describing the data, and a legend that defined the ranges of time needed to access the providers. 

Artificial Intelligence - In the last week, and in the limited amount of time I had for developing this project, I chose to focus on finalizing the writing for the summer. I described the maps, therefore the results, of the project by relating them to my beginning theoretical framework and hypothesis. In writing, I understood the limitations that I had in the project, whether they were resource-, time-, data- or information-related. Furthermore, the goal for the project was to develop a cohesive and clear form of research that focuses in community and provides results that prove useful to aid in HIV prevention endeavors. I sought to find ways in which Artificial Intelligence can be used in order to densify the amount of information and data through machine learning. In that matter, the accuracy of the project would increase and could be easily malleable if provided different information. In collaboration with AI, this research could be applied to different cities as long as a set amount of determinants were provided. In the near future, I seek to develop this project so that AI can help do cluster studies. These case studies, or digital simulations, can be used as digital experiments in cases of random HIV case increase in city clusters. These simulations would have the possibility to find weakness in the transportation system that might be unnoticeable by the maps created with ARCGIS and the manual amount of data provided.