- From: Gannon Dick <gannon_dick@yahoo.com>
- Date: Fri, 13 Mar 2015 10:08:12 -0700
- To: public-lod@w3.org, Kingsley Idehen <kidehen@openlinksw.com>
SQL queries are a semantics killer ***for those situations where the uncertainty inherent in semantics (Bayesian Interference) can not be eliminated and the deterministic solution ("frequentist Interference") is unavailable***. The terms Bayesian Interference and "frequentist Interference" are defined and conceptually validated here:(http://en.wikipedia.org/wiki/German_tank_problem). SPARQL queries preserve semantics by sampling the as yet unrealized deterministic solution. In the case where the deterministic solution is known (http://en.wikipedia.org/wiki/Birthday_problem), the apparently conflicting results are due to mere addition of quantum states. The difference in "problems" is a difference in conceptual motivation (initial conditions). This is a trivial solution to chaos effects on probability calculations, which involve a rounding at some tolerance of the constant sqrt(2*PI). Kingsley is right. To presume the infallibility of SPARQL is an ethical hazard. To presume that a DBMS is complete when it is not is also an ethical hazard. Let the debate continue, if it ends we all lose. --Gannon -------------------------------------------- On Fri, 3/13/15, Kingsley Idehen <kidehen@openlinksw.com> wrote: Subject: Re: Looking for pedagogically useful data sets To: public-lod@w3.org Date: Friday, March 13, 2015, 9:05 AM On 3/12/15 5:38 PM, Paul Houle wrote: > The goal is to show that you can do the same things you do with a > relational database, and maybe *just* a little bit more. Every RDF store is a relational database management system (RDBMS). As you know, an RDF compliant RDBMS simply group sets of RDF 3-tuples by statement predicate. We can't continue to concede the notion of a relational database management to SQL relational database management systems (sets of n-tuples grouped by Table Name). Maybe we should start referring to SPARQL compliant RDF stores as SPARQL Relational Database Management Systems, just like SQL Relational Database Management Systems which have now become synonymous with Relational Database Management System. Then "just a little more" becomes much closer to demonstrable reconciliation of "the truth, the whole truth, and nothing but the truth, in regards to relations, databases, and database management systems" :) ACID has nothing to do with what constitutes an RDBMS either, that's an a useful, but optional feature of any RDBMS. So don't fall for that baloney laden push-back when taking the SPARQL RDBMS position. We MUST end the SQL RDBMS power-grab! It has done a major disservice to the entire DBMS industry, over the last 40+ years. You have a multi-billion dollar industry that's fundamentally about companies and individuals that are data-access-heavy and data-exploitation-challenged i.e., they have tons of data ("Big Data" these days), but still can't achieve basic agility goals in regards to: accessing, integrating, and moving data effectively to the right people, at the right time, in the right form, and in appropriate context etc.. Links: [1] http://bit.ly/spasql-sql-querying-based-on-sparql-table-relation -- demonstrating that relations are relations (even when the underlying tuple organizations vary e.g., when organized as sql relational tables or rdf statements graphs) . [2] http://www.openlinksw.com/c/9C5DNHYW -- Relation . [3] http://www.openlinksw.com/c/9BVTLIAG -- SQL Relation . [4] http://www.openlinksw.com/c/9BH3NH7S -- RDF Relation. [5] http://www.openlinksw.com/c/9BDLVDX3 -- Differentiating "Database" (a Document comprised of sets of Relations [Data] ) from "Database Management System" (software for indexing and querying culled from Database Documents). -- Regards, Kingsley Idehen Founder & CEO OpenLink Software Company Web: http://www.openlinksw.com Personal Weblog 1: http://kidehen.blogspot.com Personal Weblog 2: http://www.openlinksw.com/blog/~kidehen Twitter Profile: https://twitter.com/kidehen Google+ Profile: https://plus.google.com/+KingsleyIdehen/about LinkedIn Profile: http://www.linkedin.com/in/kidehen Personal WebID: http://kingsley.idehen.net/dataspace/person/kidehen#this
Received on Friday, 13 March 2015 17:08:44 UTC