GITS: A Graph-Indexed-Tensor Structure for the Adaptive Associative-Semantic Tagging of Digital Documents

Cognitive Aspects of 2D Content Integration and Management in 3D Virtual Reality Spaces

“…Based on a literature review in [15], we have shown that an overwhelming majority of current solutions towards semantic document retrieval depend on predetermined semantic connections, either explicitly defined or extracted from human-created datasets. The camera-object independence dilemma is generated by the fact that conceptually for the user, the camera and the object should be independent (allowing users to change their viewpoint in order to ascertain whether the object being manipulated is properly aligned with its surroundings), whereas the use of the gizmo tethers the camera to the object and the 3D gizmo, making it cumbersome to change the virtual camera orientation and to access the gizmo again.…”

“…In [15], we proposed a graph-tensor representation based document retrieval model, referred to as the Graph-Indexed Tensor Structure (GITS), that is pragmatic in the sense that it does not make any assumptions with respect to semantic relationships defined a priori; instead, associations are created as the model is used.…”

“…To automatically determine the degree to which syntactic saturation can be a problem in a specific GITS model, we have proposed a Shannon entropy-based detection method in [15]. According to this method, the Shannon-entropy of a node at any given layer (i.e., corresponding to any given prefix) in the index graph can be computed based on the information contained in the next character that is typed with respect to the resulting content set:…”