Generalizing parcel data

Raw data -> google-refine for clustering and filtering -> join refined data to spatial data -> dissolve parcels using refined subdivision attribute -> buffer to fill in ROW's --> copy buffer to original dissolve layer --> dissolve again on subdivision attribute --> setup topology does not have gaps with tolerance of ~1foot --> validate topology (removes donut hole slivers)--> buffer negative value equal to the original buffer to remove overlap of adjacent subdivision polygons --> minor edits to clean up the data.