Latent Value Query: The Cutoff Grade of Data

Abstract

Organizations keep far more data than they use, and most of it is a cost. The field that set out to find value in data named itself after mining, and then noted that the name did not say what was being mined. This paper takes the comparison at its word. Mining has a precise answer to the question of what is worth recovering: the cutoff grade, the line at which the cost of recovery equals the value recovered. Applied to records, it yields a thesis: data is a liability until its cutoff grade is established. Past that point it is an asset, latent at first, waiting for its value to be found and established, and then recovered. Latent Value Query (LVQ) is a practice for lowering that cutoff: reducing the cost of turning held records into usable information, so that more of an archive crosses from liability to asset. The paper sets out the cutoff grade for data and its net-value equation; a crosswalk of mining terms with their meanings for records; the principle that a record is neither ore nor waste until a question is run against it; the two trains of work, on the material and on the question; and a set of measures for grade and recovery. LVQ is a practice of Generative Temporal Referencing, the science of placing records on a single line through time. This paper is a concept disclosure; it withholds the implementation. It has not been peer reviewed.