Intelligent Searching using Sentence Context

Fusion centers have access to terra bytes of information from both businesses and federal, state and local governments. The information ranges from computer generated databases to collections of notes with transcript of interviews performed by law enforcement personnel. Searching notes and transcripts is difficult and time consuming because humans do not use a comment set of phrases. Phrases vary due to past experiences, origin of birth and generational differences. Search engines try to compensate for these differences by performing context searches. Context searches replace specific words in the search request with other predetermined words. One can reduce false positives with an intelligent search based on grammar and English sentence structure. Intelligent sentence searching converts the each document into a set of simple sentences using only words in the predefined dictionary. These simple sentences capture the essence of the document. The conversion methodology uses synonyms, idiomatic expressions, grammar, patterns of speech and word location to create a searchable index. Because of the limited dictionary and elimination of most ambiguities, searches can be free of false positives. This paper describes the sentence context methodology, examples, and test results for a representative law enforcement report.