Using Profiles in GOSSIP to Examine Concepts Associated with the Afghan National Police in Open Source Media
Abstract
The Graphical Overview of the Social and Semantic Interactions of People (GOSSIP) is a software tool developed by Defence Research and Development Canada Toronto (DRDC Toronto). The program is designed to help the operator/analyst develop a fast and accurate impression of the relationships among entities (people, places, organizations) discussed in document collections that are too large to read through in a reasonable amount of time. Over the past few years, North Atlantic Treaty Organization (NATO) countries have developed and delivered a training program for Afghan National Police (ANP) members. The ANP are considered by some to be unprofessional, inept, corrupt, and as having little positive effect on the local population's sense of security in their communities. These qualities of the ANP were explored using GOSSIP by examining an open source media collection prepared for analysts and commanders in Kandahar Airfield (KAF). We found that when the ANP was discussed in the media updates provided to Task Force Kandahar (TFK) commanders, it was very often in a positive way. In particular, discussion about NATO s role in professionalizing the ANP dominated articles about the ANP. We propose that the extent to which the ANP is discussed in positively toned articles might lead the target audience for these articles to have an unduly positive impression of the ANP. GOSSIP is a prototype. It needs to be developed further to allow it to be used as a web-based device on a network. Future work should also enhance the tool by providing it with the ability to scrape information from various sources without the user having to load documents manually.
Document Details
- Document Type
- Technical Report
- Publication Date
- Oct 01, 2011
- Accession Number
- ADA555520
Entities
People
- Ian Lawless
- Peter Kwantes
Organizations
- Defence Research and Development Canada