Geocoding and Address Matching

Transcription

1 LAB PREP: Geocoding and Address Matching Environmental, Earth, & Ocean Science 381 -Spring

2 Geocoding The process by which spatial locations are determined using coordinate locations specified in a database table - Coordinates are expressed as X,Y values; Coordinate Systems may be Latitude/Longitude, UTM, State Plane, local, or others Results can be events [points] or routes [lines]

3 Geocoding in ArcGIS Use Geocode Addresses in ArcGIS to create features based on a table of coordinates Right-click the table in ArcMap or ArcCatalog Same tool, from menu in ArcMap

4 Geocoding in ArcGIS Toolbar Toolbox

5 Address Geocoding (aka Address Matching ) The process by which spatial locations are determined (i.e. creating points) using addresses specified in a database table Addresses in the database table are matched to address ranges specified in a reference spatial data set (e.g., streets, ZIP codes, parcels, structures) Real world coordinates are assigned to every matched address, with the real world coordinates being derived from the reference spatial data set Address-matching is an interpolative process and therefore not very precise

6 Requirements A table with addresses Example: Address: Street number and name Zone: ZIP Code or City/Town, etc A reference theme (e.g. street segments with address ranges, parcel polygons with address) Software (like ArcGIS, MapInfo, etc.) or a Web service The fields in the table correlate to the requirements of the geocoding engine

7 Process Each address is parsed into components and the components are compared with the address ranges in the reference theme When a table address falls within (i.e. matches) the address range of a street segment, an interpolation is performed to locate and assign realworld coordinates to the address (i.e., geocode it )

8 Interpolation The coordinate location of the address is estimated using the address ranges of the street segment Example: Let s say 12 Main Street is the address you want to map Segment ranges: 12 Main Street placed here (but could be to the left or right of this in reality) MAIN STREET MAIN STREET From/To Left: From/To Right: MAIN STREET

9 Interpolation Warning: check your addresses. There may not be a 12 Main Street.

10 Interpolation Warning: Actual location may be far from interpolated point location. 100 Morrissey Blvd. (geocoded point location)

11 Better Reference Data Use reference data with specific address information that does not require interpolation. Parcels Structures Center Entrance Access 32 Glenwood Ave. Parcel centroid Building centroid Building entrance Building access point

12 Address Locators The first step involves setting up an Address Locator ( GeoCoding Service in ArcGIS 8.x) An Address Locator contains the following: path and file name of the reference dataset rules for standardized descriptions of numbers, names, directions, text, etc. in the reference dataset parameters for reading address data parameters for matching address data parameters for creating output datasets Address Locators appear inside folders or at the top level of geodatabases

13 Some Address Locator Styles Style Reference Theme Type US Streets with Zone Lines US Streets Lines US Single Range with Zone Lines US Single Range Lines US Single House with Zone Points or Polygons US Single House Points or Polygons ZIP + 4 Points ZIP + 4 Range Points or Polygons Composite ** Combines existing Locators ** = Not available in SDE

14 User Steps -1 Create table of addresses and standardize the address by parsing into component parts based on the specified address style Examine the reference dataset attributes for components that are the same or similar to the standardized address

15 User Steps -2 Create Address Locator (if necessary) Adjust Geocoding Options

16 User Steps -3 Perform the matching in batch mode Evaluate results and interactively match individual unmatched records Or adjust options (in next slide) and redo batch match See ArcGIS Help or

17 Geocoding Options Spelling Sensitivity how much variation to allow in the matching process low levels would allow Main, Mane and Maine as potential candidates for 244 Main St Score score from 0 to 100 for each potential candidate based on matching all address components based on the chosen geocoding style Minimum Candidate Score threshold score from 0 to 100 that allows reporting of potential candidates Minimum Match Score rigor specification determining how well address in the table match reference layer components a score of 75 to 100 is a good match

18 Output Options Side Offset offsets a matched address so many map units to the left or right of the street segment, based on the parity of the match End Offset applies a squeeze factor at ends of line segment to prevent points from being confused with matches in other street segments X,Y Coordinates adds X and Y coordinates in output layer fields Standardized Address adds the standardized address in an output layer field Reference Data ID adds the value of the ID of the reference dataset segment in an output layer field [-1 if unmatched] Percent Along output layer field containing percent distance along reference dataset feature from the from node

19 Data Sources Addresses: Company databases, mailing lists Compile research, web or paper yellow pages Purchase Reference Data: Buy it (NavTeq, TeleAtlas, etc.) Download it free (census.gov, MassGIS, etc.) StreetMap USA extension to ArcGIS Create it or pay to create it (from parcel maps, street maps) MGISDATA.CENSUS2000TIGER_ARC in SDE in lab Can also use Alternate Name table and an Alias table

20 Geocoding Applications A company can match customer addresses against a street network of their distributor service areas in order to know which distributor will handle a customer. The information could also be used to perform marketing analysis for each service area A police department could use address geocoding to analyze crimes by address. These locations can be mapped and analyzed with other demographic data Emergency dispatch operators could use geocoding to enter an address, determine who should respond, and route emergency vehicles and personnel to the address

21 Geocoding Applications A school district could use geocoding to match student addresses against a street map of a city. Once the homes of students are located, school assignments and busing plans can be created and analyzed Environmental engineers can identify potential impacts of storage facilities for hazardous materials in populated areas by matching the site addresses to a theme containing census data Data development Create layers like police and fire stations, schools, town halls, libraries, prisons

22 Caution, Warnings Address matching can be a laborious process and is a function of the accuracy of the address table and reference theme. Use software s batch matching functionality, then match interactively for those not matched An address table may have typos, incorrectly spelled street names, blanks, common names, etc., and data can be out-of-date. Some buildings can have vanity addresses (e.g. One Executive Plaza )

23 Caution, Warnings Make sure and be prepared to have a verification process. i.e. a good atlas The reference data may not be spatially accurate at large scales (like TIGER roads), so resulting points may need to be adjusted using better basemap layers (like ortho imagery)

24 Help The Geocoding Tutorial coding_tutorial/ /