• No results found

GIS Procedural Guide Geocoding / Address Matching in ArcGIS Created by Steve Zuppa, Map Library Assistant - Serge A. Sauer Map Library, 2008

N/A
N/A
Protected

Academic year: 2021

Share "GIS Procedural Guide Geocoding / Address Matching in ArcGIS Created by Steve Zuppa, Map Library Assistant - Serge A. Sauer Map Library, 2008"

Copied!
5
0
0

Loading.... (view fulltext now)

Full text

(1)

GIS Procedural Guide

Geocoding / Address Matching in ArcGIS

Created by Steve Zuppa, Map Library Assistant - Serge A. Sauer Map Library, 2008

Introduction

In a traditional business database, street addresses are the most common form of geographic data. Using a GIS, this geographic component can be estimated and mapped based on an existing spatial database, such as a street network. This process is called ‘address matching’ or ‘geocoding’.

Step 1 – Find the Data

You need at least two data elements to geocode in ArcGIS: a table containing the addresses you would like to map and a proper road network data file.

a) Address Table

You will need to build a table consisting of addresses you would like to geocode. Normally some type of description of the address (ie. business name, or zoning type etc.) should be present in the table as well.

Your table can be saved in a number of different formats (*.xls, *.csv, *.dbf, etc.),

depending on the software installed on your computer. All UWO computers have access to Microsoft Office, so for this example, Excel has been utilized to create an address table for Record Stores in London, and has been saved in standard xls file format.

Data was gathered from www.Canada411.ca by undertaking a simple search for “Music-Records, Compact Discs & Tapes” within London, Ontario. Our table contains columns for Business ID, Business Name, Address, City Province, and Postal Code, however only the Address field is truly necessary (see below). Notice that the addresses are stored within one field in a very simple format (number name suffix).

b) Road Network File

(2)

For restricted (UWO only) data, see the DMTI CanMap RouteLogistics package

http://geography.uwo.ca/maplibrary/geospatial/dmti_rl_data.htm available through the Map Library or the Data Resources Library.

For this example, we will use Statistic Canada’s 2008 Ontario Road Network file. Since we are only looking at addresses within London, the Road Network file has been clipped to London city boundary.

Step 2 – Build and Address Locator

Open ArcCatalog and use the Catalog Tree to navigate to the folder where your data is stored. Note: You may need to Connect to Folder if you cannot see your folder in the Tree.

Right click on folder name and select New ► Address Locator. Scroll down to select the US Streets (File) option (this is just a generic name).

In the following window, give your new Address Locator a suitable name and in the Primary table tab select the 2008 Road Network as your reference data. You may want to check “Store relative path names” if you plan on moving your data to another folder or computer Make sure the appropriate Fields are selected below. If you are using the Statistics Canada Road Network File, the matching Fields are:

House From Left: ADDR_FM_LE House To Left: ADDR_TO_LE House From Right: ADDR_FM_RG House To Right: ADDR_TO_RG Prefix Direction: <None> Prefix Type: <None> Street Name: NAME Street Type: TYPE

Street Direction: DIRECTION

The other options allow you to adjust matching scores and whether you want your point offset from the road line. We’ll leave it as the default. Your window should look similar to the graphic below. Click OK.

(3)

Step 3 – Geocode Addresses

Open ArcMap and add your Road Network file. Next, add the table containing the addresses you would like to geocode.

*Note: if you are using an XLS file, you will need to double click on the file name in the Add Data window, and then select the appropriate sheet (in this case it would be

record_stores$).

Make sure the Source tab is selected at the bottom of the ArcMap table of contents, so you are able to see both the road layer and the address table. See graphic below.

Right click on your table name and select Geocode Addresses. You will need to Add the Address Locator you created in the previous step. Click on Add… and browse to the folder where you stored your Locator. Click Add and then OK.

(4)

The next window will indicate how many of your address points were matched and unmatched. If all of your points matched with a suitable score, then you don’t have to worry about the next step. If you have any that didn’t match for some reason, you can attempt to manually match the address to your data.

In this example, all of the records stores but one matched successfully. Now we’ll try to figure out why the one address didn’t match. Click Match Interactively.

In our example, the one address that didn’t match was for Speed City Records. The address we have recorded for this record store is 2909 SPRINGBANK DRIVE.

Looking through the Interactive Review folder and scrolling through the list of potential matches (Candidates) and concentrating on the numbers shown in the LeftFrom, LeftTo, etc. it looks like there are no numbers up in the 2000-3000 range. All the potential candidates have a relatively low score of 38% or less. This could indicate a problem with the table address data.

Checking back with www.Canada411.ca reveals that the actual address for Speed City Records is actually 299 SPRINGBANK DRIVE. It seems there was a typo in the data entry phase of our geocoding project. By clicking the Modify button and changing the HouseNum field to 299, we now have a candidate that has a match score of 100%. We’ll select that candidate and click the Match and then the Done button. Click done in the Review/Rematch Address window and your points should be visible in the ArcMap display.

You now have successfully Geocoded your addresses! You can use your point shapefile layer in any way you see fit. Here’s a sample map made with some of the record store points and free Canada-wide NTDB data (available for download via Geogratis, see

http://geogratis.cgdi.gc.ca/geogratis/en/product/search.do?id=8147):

(5)

" " " " " "" " " " " " " " " " " " " " " " " " " " " " " " " " " " " " " " " " " " " " " " " " " " " " " " " " " " " " " " " " " " " " " " " " " " " DUNDAS ST DUNDAS ST RIC HM ON D ST YORK ST WE LLIN GTON ST HORTON ST E QUEENS AV E STANLEY S T Grooves Dundas Discs Mandala Books Sunrise Records Music & Video Place

L'Atelier Grigorian

City Lights Bookshop

References

Related documents

(j) Back in the layer properties window you can select pre-packaged color schemes (called color ramps) or click on each color box to select new colors.. (k) For the sake of

As a woman ages past 40, the uterine lining can begin grow inside the muscular wall of the uterus (adenomyosis) and cause pain either all month long or in cycles.. This is

Even so, the insurance company may deny your claim, so we encourage our clients to return to work as soon as they’re able and medically cleared after an auto accident.... Know

1 EXECUTIVE SUMMARY: A member’s start-up of a local business in conjunction with the starting of a Church of Christ can be a marriage made for Heaven.. Each can

2 reception rooms • study and utility room • kitchen/ breakfast room • 4 bedrooms and 4 bathrooms • indoor hydrotherapy pool • detached two bedroom cottage •

Thalamic nuclei segmentation using thalamo-cortical projections (Behrens et al. 2003 ) used seven cortical targets to draw probability distribution of connections from voxels within

Email filtering is a process of monitoring incoming email and then taking the appropriate action to protect against Spam and viruses. Certain criteria are set to determine if an