This is the Github Repository for the Knowledge Graph Conference Book Club around the book " Designing and Building Enterprise Knowledge Graphs: Synthesis Lectures on Data, Semantics, and Knowledge" with Juan Sequeda and Ora Lassila
3
stars
0
forks
source link
how to capture reasoning in phase-1 knowledge report of iterative KG-building process? #21
Where might symbolic and statistical reasoning fit in to the "knowledge report" framework of Chapter 4, e.g. in Tables 4.2 through 4.4?
It seems that some decisions re: reasoning would directly impact trust in answers to business questions. For example, questions about "family members" (is a step-child's spouse included?) or about "outliers" (> 3 sigma for a fit normal distribution?) may rely on consistent symbolic and non-symbolic (statistical) characterizations, respectively, for which a C-suite report-reader would not want different data consumers, e.g. analysts, to produce different values based on the same data product, i.e. based on the same KG projection.
Where might symbolic and statistical reasoning fit in to the "knowledge report" framework of Chapter 4, e.g. in Tables 4.2 through 4.4?
It seems that some decisions re: reasoning would directly impact trust in answers to business questions. For example, questions about "family members" (is a step-child's spouse included?) or about "outliers" (> 3 sigma for a fit normal distribution?) may rely on consistent symbolic and non-symbolic (statistical) characterizations, respectively, for which a C-suite report-reader would not want different data consumers, e.g. analysts, to produce different values based on the same data product, i.e. based on the same KG projection.