‘Context’ defined (as cited from the Merriam-Webster Online Dictionary): 1: the parts of a discourse that surround a word or passage and can throw light on its meaning; and, 2: the interrelated conditions in which something exists or occurs: environment, setting.
While context is clearly important, today it is consistently traded for brevity. We communicate in half-sentences (got it), atypical abbreviations (k), emoticons (J), etc. However, while striving to maximize efficiency, we provide the minimal amount of information possible and hope or assume that a full understanding is received. When a full understanding is not received or misinterpreted due to a random flow of textual shortcuts within a narrative (or unstructured data), we fill the gaps with our own context – which may provide a completely erroneous reading of intent.
Communicating electronically within law enforcement agencies is no different --- officers readily add important narrative in code, abbreviations, or geographic-specific nomenclature. Given these challenges, the unstructured data often lends minimal value and adds little context to the structured data housed in an agency’s Computer Aided Dispatch and/or Records Management System. For example, the narrative may include descriptors of items found in or around a scene that may prove helpful to another investigation; being able to identify items of interest and automate a search process could be invaluable.
We are working with law enforcement to harness its unstructured data via text analytics – applying automated extraction and interpretation to existing free text will help agencies maximize its data value and shape response, tactics and strategies. Text analytics provides law enforcement the promise to contextualize its known, useful, structured data within the officer’s narrative – leveraging the on-the-ground expertise and experience to better understand a given occurrence or situation.
After all, it’s all about context.