User Tools

Site Tools


Meta-Geocoding

Introduction

The Meta-Geocoding module from Spatialytics ETL is a tool used to geocode data by using existing external geocoding engines (via exposed Web Services / APIs). It can use many geocoding engines - Open source or not, free or not - that are known on the market. Among these engines are:

  • Google,
  • Yahoo! Placefinder,
  • Bing
  • OpenAdress,
  • GeoNames,
  • OpenStreetMap (Nominatim),
  • Wikimapia,
  • Semantic MediaWiki.

The engines listed above allow the Meta-Geocoding module to offer geocoding and reverse geocoding capabilities. Geocoding is the action of attributing geographical coordinates (longitude/latitude) to an address. Reverse geocoding is the opposite.

The Meta-Geocoding module takes in input text data that can be transformed using Spatialytics ETL steps and gives as output a point type geometry. Accordingly, reverse geocoding takes points as input and gives addresses as output.

The Meta-Geocoding module is a powerful tool thanks to the many geocoding engines it uses. Therefore, the use of these engines implies a certain knowledge of their possibilities and an acceptation of their terms of use.

Installation

Here are the installation instructions for the Meta Geocoding Module:

  • Download and extract the Meta-Geocoding module archive. The following directories will appear:
  • Plugins : Contains the module itself organised as sub-directories containing steps/GeocodingPlugin
  • libext : Contains the Java librairies used by the module.
  • Copy and paste the archive content in the <SETL_HOME> directory.

Then launch Spatialytics ETL and the Meta-Geocoding module will appear under the name Geocoding Module in the Transform section. You can then use it as any other Spatialytics ETL step.

Configuring the module

The use of the Meta-Geocoding module is the same as any other step. Follow these instructions to configure it.

  • In Design mode, from the Spoon interface, drag and drop the module.
  • Double-click the module to open the configuration window.

Section A: Geocoding Name and Type

Step Name: This is where the step name is typed

Geocoding Type : This is where the user can choose between geocoding or reverse geocoding.

Section B: Service Configuration

Geocoding Service : The user can choose the geocoding engine here. It can be Google, Yahoo, OpenAddress, etc..

View licence: This is a mandatory action for the user to accept the terms of use.

Service URL: Used to type in the URL of the geocoding engine. For most of the engines except Semantic Media Wiki, the user does not have to specify the URL as it is automaticaly entered.

Filter by: Allows the geocoding filtering by administrative regions, airport, country, intersection, municipality, etc. Unfortunately, this function is not supported by all engine.

Provide API key : Use this button to type in the API key of the choosen engine. This key is necessary to use the geocoding services of the engine.

Section C : Autres options

Get first match row only: If this box is not checked, all the results of the geocoding of a specific address are sent back. If it is checked, only the first result is sent back.

Field to geocode: Specifies the column to be used for geocoding or reverse geocoding. For a geocoding operation, the column must have an alphanumeric type. For a reverse geocoding operation, the column must have a geometry type.

Result field name : Specifies the column where the result of the geocoding will be stored.

Add full address from engine: When this box is checked, the Meta Geocoding module returns the full address that the engine used for geocoding. This allows the user to update his database with this address. Note that this option is not available for a reverse geocoding operation.

Full address field name : When the Add full address from engine box is checked, this field specifies the field name of the full address returned from the engine.

Example

See the following transformation:

The Meta-Geocoding module receives data from a file.

Step 1: Configure the Meta Geocoding module

In the figure above, we can find the following configuration:

  1. This is a geocoding operation
  2. It will use the Google engine
  3. The terms of use have been accepted
  4. No filter will be used
  5. The user provided an API key
  6. The user chose to get all points returned for the address (type 1:n geocoding)
  7. The input column containing the addresses to geocode is ADDRESS
  8. The output column that will contain the geocoding results is RESULT
  9. The user wishes to get the exact address used to geocode from the engine
  10. The column name that will contain the address returned by the engine is DESCRIPTION

Step 2: Preview the result

  • The first column, named RESULT, is the output geometry from the Meta-Geocoding module. Note that the geometry type is POINT.
  • The second column, named GEOCODINGQUALITY, gives the quality level of the geocoding or reverse geocoding. This information depends on the engine used for the geocoding. Not all geocoding engine return this information.

For the Google geocoding engine, please refer to their geocoding API documentation at https://developers.google.com/maps/documentation/geocoding/#Results.

For the Yahoo! geocoding engine, please refer to their geocoding API documentation at http://developer.yahoo.com/boss/geo/docs/supported_responses.html#address-quality.

  • The third column, named DESC is the address returned by the engine corresponding to the geocoding result.
en/spatialytics_etl/006_modules/meta_geocoding.txt · Last modified: 2013/02/28 06:21 by lvaillancourt

Page Tools