A new Role for Geo-location evidence in eDiscovery

Posted on March 28, 2011


The surge and ubiquity of geo-based technology, such as GPS apps in smart phones, is giving rise to a new world of location based electronic discovery (LBED). The geo-spatial data stored within a custodian’s devices (cell phones, iPads, etc.) and ESI metadata may conflict with their own reports about their actions or whereabouts at a specific time.

These days, it is also possible to use text mining technology to derive location based information in textual documents with high precision. Subsequently, geo-mapping technology enables legal professionals to map documents to geographic locations. The result is a graphical view of potentially relevant patterns in the evidence, events and sites.

Text mining technology can automatically recognize the names of towns, cities, regions, countries or even detailed addresses. The intelligent software then considers the context of the document in order to determine what the semantic role of the address is: is it someone’s home address? Is it the location where a meeting will take place? By using the proper linguistic technology to obtain as much understanding as possible, one can determine with high accuracy the location of events such as meetings, crimes, travel patterns, and border crossings.

When location is relevant or disputed, that type of proof can be preserved, collected, and submitted to the court—just like the geo-spatial information stored within a person’s mobile device.

This type of technology has already been applied in the real world: for example, the European Union Police Mission in Kosovo has been using ZyLAB software to prove that the routes taken by military troops directly correlated with the rural villages where crimes are being investigated. The following satellite image shows part of Kosovo with some sample documents linked to specific geographical locations (this is not real evidence material; this is just to show the concept).

The applications for this technology extend far and wide. Consider how organizations could extract text to pinpoint branch offices which may be engaged in activity that conflicts with various regulations such as the upcoming UK bribery act and the US Foreign Corruption Practice Act.

Another example is the website http://twittermood.org/ , where the sentiment (positive or negative) of Tweets is mapped on the geographic location of the Tweeter. You can imagine how sociologists could study the state of happiness of populations based on an unending supply of empirical data.

The automatic recognition of geographical location in textual information can also be used to redact personally identifiable information. The following example shows how an address and phone number is automatically redacted before a document is disclosed under a FOIA request.


 The future will show us how digital footprints can belie an alibi and how geo-location data and satellite maps will change the litigation and information management landscape.  But to me it is clear that this new technology will bring a completely new dimension to eDiscovery and investigations and bring fact-finding to a new level!