Neighborhoods are organic entities that are in a state of constant change and are driven by the specific context of the problem being investigated. The subsequent lack of consensus on a universal geographic definition for what constitutes a neighborhood can lead to biased interpretations of relationships between human activities and place. Further, while existing geographical information system software allows users to combine a range of geographic objects to generate regional units of analyses, their design does not explicitly assess how changing patterns, such as populations, impact the data expressed within them. This research develops an exploratory geographical information system framework that allows users to dynamically delineate neighborhoods based on user-specified characteristics. These include socioeconomic and similar measurements of neighborhood classification from information obtained from secondary data sources, including parcel data, land use/land cover information, and attribute data provided by the United States Postal Service. The proposed methodology creates custom geographies from readily available tract data obtained from various federal and state data repositories to produce indices. By allowing the user to dynamically weigh the combinations of variables used to define their neighborhood, this thesis introduces a solution to a common analytical problem in the discipline.