The Life Cycle of Trusted Data:
From Acquisition to Persistence... Or Not
Continuing from the previous webinar on the importance of trust in knowledge work, the categories of trustworthiness, and how apparent data quality impacts perceptions of trustworthiness, this webinar carries the trust concepts through into analytics and data management.
We begin by identifying facets of trustworthiness which are mathematically related to data uncertainty and introduce paths to quantification of those values. These are things like obtaining the catalog drift values to use on uncalibrated sensors or finding r-squared in the original papers defining common computations using empirical correlations (like Archie's Law for water saturation).
We then discuss how to conveniently represent trustworthiness as metadata and carry that information along as data makes its way through the seismic to simulation (or exploration to abandonment or whatever a company uses) workflow.
We end with a discussion of the data management aspects of trustworthiness. Does trustworthiness become stale over time as data ages? How would we persist trustworthiness scores in a database? Do we need to persist them or are they derivable into the future?