Paper
9 April 2007 Semantic search via concept annealing
Author Affiliations +
Abstract
Annealing, in metallurgy and materials science, is a heat treatment wherein the microstructure of a material is altered, causing changes in its properties such as strength and hardness. We define concept annealing as a lexical, syntactic, and semantic expansion capability (the removal of defects and the internal stresses that cause term- and phrase-based search failure) coupled with a directed contraction capability (semantically-related terms, queries, and concepts nucleate and grow to replace those originally deformed by internal stresses). These two capabilities are tied together in a control loop mediated by the information retrieval precision and recall metrics coupled with intuition provided by the operator. The specific representations developed have been targeted at facilitating highly efficient and effective semantic indexing and searching. This new generation of Find capability enables additional processing (i.e. all-source tracking, relationship extraction, and total system resource management) at rates, precisions, and accuracies previously considered infeasible. In a recent experiment, an order magnitude reduction in time to actionable intelligence and nearly three orderss magnitude reduction in false alarm rate was achieved.
© (2007) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Kirk A. Dunkelberger "Semantic search via concept annealing", Proc. SPIE 6570, Data Mining, Intrusion Detection, Information Assurance, and Data Networks Security 2007, 65700K (9 April 2007); https://doi.org/10.1117/12.720039
Advertisement
Advertisement
RIGHTS & PERMISSIONS
Get copyright permission  Get copyright permission on Copyright Marketplace
KEYWORDS
Annealing

Detection and tracking algorithms

Heat treatments

Databases

Genetic algorithms

Materials science

Statistical analysis

Back to Top