Easy Georeferencer might not be the best suited for your needs, so on this page I have included information about other alternative geocoding options available on the internet and some of their pros and cons. For a more exhaustive list of options (there are tons, but only some are easily usable for non-technical users), see the excellent list of georeferencing tools by the Texas A&M University’s Geography Department.
Types of Geocoders Mentioned Here
- Desktop Software
- Commercial Services
There are many websites that lets the user georeference directly on a website. Some of these are both easy to use and accurate. BatchGeo is one such website that allows the user to copy and paste a selection of rows from a dataset to the website window, which populates a Google Map with point-markers and the possibility to download the point data in .kml format.8 Though it is an incredibly easy to use service, the free version only allows 250 data points at a time, and the payable Pro version only allows 15 000 data points. Such size limitations put serious dampers on using this service to geocode large social science datasets. Furthermore, even if the desired dataset is within the BatchGeo limit, the downloadable .kml spatial version of your data is not immediately usable for further analysis in GIS software and may need to be converted before use.
GeoCommons is another user-friendly website that can georeference data, one that seems to have several advantages. GeoCommons is not just a mapping website, it has many additional features such as spatial analysis and time-animation. Its geocoding feature is based on GeoNames coordinates data and requires that you sign up for and log into your free GeoCommons account. The website allows you to actually upload your data file instead of having to copy and paste, and allows you to download the georeferenced output file in the popular shapefile format used by most GIS software and thus allowing for immediate analysis. Unfortunately, there is still a georeferencing limit of 5000 data points for each upload.
Other services are reliant on having a Google account. One is the MapAList website intended for social media sharing that only processes online Google spreadsheets and lets you download the limited kml format. Another service comes directly from your Google account by uploading and importing one’s Excel sheet into a Google Fusion Table (max 250 MB per account), after which one only needs to click “Visualize” and “Map” to process and see the results, with the option to download it in the kml format. Of course, the limit with these approaches is signing up for a Google account if you don’t already have one. More importantly, it is also likely that these Google-dependent services are subject to the Google API’s processing limit of 2500 records per day.
There are also some other free websites that have simpler and more limited functionality and website designs. These websites are based on copying and pasting your datapoints into the webpage, and relies on all geographic names being in one and the same field (e.g. “street-city-province-country”). Instead of returning spatial files these websites only return a text box with the coordinates of your data, which you have to copy, paste, and convert to a spatial file format before you can do further analysis. One website, FindLatitudeAndLongitude, seems to conduct fast geocoding but is limited to “a few hundred addresses” at a time. Stephen P. Morse’s Batch Conversions of Address to Latitude/Longitude (Forward Geocoding) appears to have no theoretical limit, but is not very fast and is likely to time-out for very large copy-pastes.
The recurring theme that we have seen in all of the georeferencing websites reviewed here is that they often put usage-limits on their users to help save bandwidth or entice people to buy their service. Even if these limits were overcome there would always be the limitation inherent in any website solutions: the vulnerability from and dependency on a stable internet connection, that it takes time to upload large datasets, and that waiting for the website to georeference very large datasets and provide you with a download link may take many hours or even overnight—all of which brings up the potential problem of your internet browser session timing out. It is probably no coincidence that these websites seem to primarily market themselves for non-profit and mainstream social media-use whose georeferencing-needs may be less demanding than for academic use.
Next to consider are the georeferencing abilities of desktop georeferencing softwares. One of these is the GeoLocate software. Geared specifically for georeferencing natural history collections, this software comes packages with reference data and does its georeferencing offline, and even allows geocoding of obscure locations like (5 miles northwest of X location). The backdraw is that its graphical interface seems like it comes from a different era and is not entirely intuitive to use. And for georeferencing international locations one has to download an additional extension pack.
Another one is the Texas A&M Geoservices desktop software. Created by a University center that specializes in georeferencing, this software loads your dataset from your machine and geocodes it over the internet via their own webservice. Although the center has a lot of know-how and provide many useful geocoding resources for non-technical users, their desktop software still has some backdraws in terms of its user-friendliness (as they themselves warn their users about). Their software has a limit of 2500 records per day, does not accept Excel files, and requires signing up for a user account and performing some technical tasks like installing third-party Mircrosoft database drivers and inputting an API key (though they also have an equivalent website service where these technical steps are not required).
There are also other desktop software alternatives that are not primarily about geocoding but that nevertheless have geocoding functionality. These tend to come from corporate industry-standard softwares that cost large sums of money but that many Universities and businesses make available to their students and staff for free. The two examples I could find that allow automated georeferencing are the ArcGis mapping software, and the SAS statistical software. Note however that these still rely on having internet contact for each session and require several advanced steps of setups and preparations. For instance, in ArcGis the user is expected to add server connections and create “locator files”. Furthermore, it appears that even though ArcGis cost money to have in the first place, the batch geocoding option which used to be included in the main license package now requires an additional fee through an ArcGIS Online Organization subscription. In SAS, the instructions seem to require use of the local SAS syntax to first set up some “macro variables” and then to call on the geocoding program routine. If you possess these minimal technical skills and have access to these services through your institution then one of these is likely to be the best, fastest, and most reliable option.
Finally, there are some choices for priced geomatching. BulkGeocoder is one such service with a user-friendly webpage design that lets the user upload their dataset file and later on download the geocoded version with coordinates and accuracy fields appended in exchange for a payment option. USGeocoder (limited to the United States) will geocode your dataset if you send them your file via email and pay a fee based on the amount of records to be coded. SmartyStreets is another service that is seemingly very powerful and fast, and for most datasets (over 45 records) charges a fee. One promising caveat here though is that they claim to offer unlimited use of their service for non-profits, schools, religious organizations, and libraries, although I am unsure if this would also apply to researchers at Universities or other research institutes. Another one that I was recently made aware of ZP4 which comes for a relatively flat rate on a DVD and geocodes to address coordinates and neighbourhood blocks, within the US. Finally, there are also several big-industry geocoding corporations like Pitney Bowes, but the prices and quality of these types of services are in a whole other league and probably a bit overkill for most academic use, so I will not mention these here.
Some academics may still find that the financial cost of buying the service outweighs the time-cost of geocoding the data themselves. However, given the fees required for each geocoding, most of these commercial services may be more usable for businesses than they are for academic research which is already heavily constrained by funding budgets.
Given the various limitations of existing methods mentioned here, it is clear that academia (and society more generally) could benefit from more and better automated georeferencing methods. It is the hope that the new freely available Easy Georeferencer software will address some of these previous limitations and provide a more suitable georeferencing tool for academic research. If this sounds interesting, you can either read more about its features, see pictures of what the program looks like, or get started downloading and using it.