The joining of geospatial datasets is required to utilize the complete set of information available in each of them. There are many open source geospatial datasets available such as GeoNames, Open Street Map, Natural Earth and to get a comprehensive dataset with the union of all available information it is important that such datasets are linked optimally without redundancy or loss of information. Many of the geolocations on digital maps are not classified for importance because of the lack of additional information such as population or administrative level. A way to give an importance scale to the names is by linking the GeoNames to other datasets (OSM, natural earth). OpenStreetMap data provides a limited number of place classifications (such as city, town, village). For the best cartographic results we need classes that are a little more comprehensive about how they rank cities. The challenges faced include geometry searching, matching, buffer determination, local regional naming text inclusion and accuracy. This has been achieved by the current research work where presently GeoNames, Natural Earth and Open Street Map data tables have been merged with the union of all their attribute columns resulting in a complete geospatial dataset with place accuracy of atleast 95 % for any given country dataset. The data tables at global level consist of hundreds of thousands of rows with each row depicting a geolocation. The geometry, name and geo-id complete and fuzzy searching and matching around a buffer of 50 km took a minimum of 30 s to maximum 1 min in a commodity computer with 2 GHz, 2 GB memory, according to size and complexity of the query run for a country which could have a list of points ranging from a dozen to several hundreds. The future aim is to ultimately do this for global datasets to create an all-encompassing geodata bank having such information as administrative, political, ecological details from important databases as GAUL, SALB, GADM etc.