Using Natural Language Processing with Geotagging

If your data includes images of text, for example scanned articles or PDFs, you can use the Natural Language Processing (NLP) Extension to enhance the accuracy of Geotagging.

Before You Begin

  • Python version 2.7 must be installed on the same location as Voyager Server. If Python is not installed, NLP will not function and Voyager will display an error.  Download Python here. If Python is not installed, NLP will not function.

  • Make sure that the NLP Service is enabled.  You may need to download the NLP Python file if you have not already done so. See Downloading the NLP Python Libraries and Configuring the NLP Extension for more information.

  • You must add the nlp worker to the Pipeline for the location you are geotagging, otherwise there will be no NLP location information present for the geotagging process

  • You may also need to re-scan or re-index the location before you begin Geotagging.

Using NLP in Geotagging

When the nlp worker has been added to the Pipeline and there is appropriate content for it process, it stores location information in the nlp_place field.

To use NLP in Geotagging:

  1. Go to Manage Voyager > Discovery > Locations 

  2. Click the Edit icon for the location you want to geotag

  3. Select the Geotag tab

  4. Select Geotag from a Standard Gazetteer (you can also use a Custom Gazetteer if you have already created one)

5. In the Geotag dialog, enter nlp_place in the Select the Text Field to Geotag box

6. Click Save

When you are done, you will need to re-scan or re-index the location to pick up the new Geotagging information. If everything is successful, you should see location information in the Details Page for an individual record.