NetOwl at ESRI Federal GIS Conference, Jan. 29-30

Geotagging

ESRI is hosting their annual Federal GIS Conference on January 29th and 30th at the Walter E. Washington Convention Center in Washington, DC.   NetOwl will be there showing how our industry-leading NetOwl Extractor product is being leveraged with ESRI, the leading GIS provider, to go beyond what’s possible with simple geospatial analysis of text content.

With its ArcGIS suite of tools which range from standalone desktop applications through enterprise scale platforms and services, ESRI is the preeminent supplier of GIS-related capabilities including maps and a variety of geospatial analytics components.  For many GIS applications, all of the data that is available already has latitude/longitude or other coordinate information associated with it.  For other key applications, however, one important source of data that typically does not have explicit latitude/longitude values is unstructured text.  That information is buried in the place names within text itself and not in any special metadata field.  Finding and disambiguating place names in text and turning them into coordinate values is where NetOwl’s Smart Geotagging comes in.

NetOwl Smart Geotagging: Advanced Place Entity Extraction and Geotagging

Whether working on open source text content or other sources of text-based documents, NetOwl Extractor can automatically extract from any unstructured text document:

  • Over 100 types of entities including the names of people, places, and organizations
  • Over 30 types of links that connect entities together
  • Over 100 events that entities may be involved in

While some other tools are able to identify a variety of specific coordinate expressions such as “41 25 01N, 120 58 57W” or “48 N 377299 1483035” in unstructured text and use that information to associate “points on a map” to the document, NetOwl’s geotagging capabilities go well beyond simple coordinate extraction and normalization.  Some of the key entity types that can be geocoded with NetOwl include:

  • Various latitude/longitude and other coordinate expressions like MGRS and UTM. NetOwl normalizes all of these expressions to decimal latlongs.
  • Named place entities like countries, provinces, cities, and districts. Conceptually, this sounds simple, but when all you know is that something happened in the city of “Springfield”, with no explicit mention of which Springfield it is, NetOwl can help disambiguate the correct location from other information present in the text.
  • Relative places. In some cases, the best information you can get about a location is a ballpark area like “10 miles NW of Baghdad”.   NetOwl can identify such relative place entities and perform the coordinate math needed to normalize such a place entity to an absolute coordinate.

Beyond Place Entities: Geotagging Events and other Types of Entities

At the geospatial entity level, NetOwl provides for a rich geotagging capability that goes beyond any other unstructured text geocoding product.  By leveraging NetOwl’s advanced Link and Event Extraction capabilities, key entities and events, which by themselves are traditionally not geocodable, can now be connected to geocodable locations and presented on a map in their proper geospatial context.

Through various “links” that NetOwl can extract, the names of people, organizations, and other entities can be put on the map at their proper location.  For example,

  • A representation of a “person” can be put on the map at a location because the text indicates they “reside” there.
  • An organization can be put on a specific location on the map because the text talks about where the organization is headquartered.

Similarly, the locations of over 100 potential extractable event types can be used to place event-indicators on the map such as:

  • Locations where a meeting is taking place
  • Locations where a bomb went off or any other conflict-related event occurred
  • Locations where one or more people travelled to or from

By leveraging the semantic information NetOwl can extract from a collection of documents, along with its geotagging capabilities, geospatial applications can display a far richer geospatial view of the key information from those documents that go well beyond a “set of points” on a map.

Visit our Booth

To learn more about our geotagging capabilities and to see it in action, stop by our booth (#823) at the Esri Federal GIS Conference!