Semantic Knowledge Representations for Soft Data Fusion - Efficient Decision Support Systems: Practice and Challenges from Current to Future - page 7

Fig. 3. Endsley's model of situation awareness (adapted from [Endsley, 1995])
Endsley's model illustrates three stages or steps of Situation Awareness formation: perception,
comprehension, and projection.
Perception (Level 1): The first step in achieving Situation Awareness is to perceive the status,
attributes, and dynamics of relevant elements in the environment. It also includes
the classification of information into understood representations and provides the basic
building blocks for comprehension and projection.
Comprehension (Level 2): Comprehension of the situation encompasses how people
combine, interpret, store, and retain information. It includes the integration of multiple
pieces of information and a determination of their relevance to the underlying goals.
Comprehension involves a synthesis of disjointed Level 1 Situation Awareness elements
through the processes of pattern recognition, interpretation, and evaluation. It includes
developing a comprehensive picture of the world, or of that portion of the world of concern
to the individual. Furthermore, as a dynamic process, comprehension must combine
new information with already existing knowledge to produce a composite picture of the
situation as it evolves.
Projection (Level 3): The third level involves the ability to project the future actions of the
elements in the environment. Level 3 is achieved through knowledge of the status and
dynamics of the elements and comprehension of the situation (Levels 1 and 2), and the
ability to make predictions based on that knowledge.
3.2 Graphical methods for soft data fusion
Graph-based structures seam to be key structures for situation understanding and high-level
information in general. Graph-based formalisms are easily readable and understandable
by humans. Furthermore, graphs are a natural way to represent several ideas or objects
interacting with each other. Therefore, information fusion based on graphs structures is a
major stake
Sambhoos et al. (2008) relates about Inexact Graph Matching for real-time Fusion. Information
items extracted from texts written in natural language are stored as RDF 1 triples. Each triple
contains a subject, an object and a predicate (or relation) that exists between the subject and
the object. The triples are considered as simple graphs.
Each text message may lead to the extraction of several RDF triples. These triples are then
organized in more complex graph structures called “observation graphs”. The larger graph is
created by linking the commonly named nodes from the initial triples.
1 Resource Description Framework (RDF) is a graph based model proposed by the World Wide Web
Consortium (W3C), used to describe web resources.