Re: The anatomy of plagiarism that was made by authors of "Anchor Modeling"
Date: Tue, 31 Mar 2015 11:06:50 -0700 (PDT)
Message-ID: <873f1c87-7085-4a46-845a-c6f8c3ca63d2_at_googlegroups.com>
In my my last post (from 26/03/2015) I noted that "Anchor Modeling" is incorrect in
fundamental matters and that it is not in accordance with fundamentals of database
theory.
In this post I will concentrate on two fundamental matter that were first solved in
my papers. Five years later, the authors of "Anchor Modeling" plagiarize these
results. Those results are:
(i) The presentation of the real world using the theory of states (ii) Decomposition of data structures into atomic structures
Theory of states
Some of the very important direct consequence of my theory of states, I wrote in this thread in my post from 17.03.2015. These consequences of my theory of states for the first time addressed the aforementioned important problems in logic, semantics, Model Theory and database theory. I mention this because I want to emphasize the great proportions of this plagiarism.
Decomposition of data structures into atomic structures
When it comes to decomposition of the data structure, then the main paper of "Anchor Modeling" from 2009 is unprecedented. In this paper the decomposition of data structure was not proven. Moreover, "decomposition" is not even mentioned in this paper. The authors of "decomposition" are only defined "Historized and Static Attribute" and "Historized and Static Tie".
"Decomposition" is what many theorists has failed to solve. Two of these unsuccessful
attempts were made by E. Codd and Date & Darwen. Codd in RM / T, and Date & Darwen in 6NF.
In order to explain the anatomy of the decomposition of data structures, I need more space.So, below I will write this explanation.
We can notice that there is no procedure that can put an arbitrary relation in 6NF. But the authors of "Anchor Modeling" are addressed their work with the following title: "Anchor modeling - an agile modeling technique using the sixth normal form for structurally and temporally evolving data". Obviously, part of the text "using sixth normal form" from this title should give readers the impression that in "Anchor Modeling" is proved "decomposition".
First, I would like to say a few words about "sixth normal form" (6NF for shortcut). This is about the the most important thing in db theory, that is, it is about the decomposition into atomic structures. It is clear that the atomic structure must be in the following format: (K, A) where K is the simple key and A is an attribute. Codd has tried to resolve the matter with the introduction of "Simple Key". But the RM / T is unsuccessful attempt. In this paper Codd did not understand some things. The most important thing that he did not realize is that there is a "History". He tried to resolve this matter by using "invisible surrogate". However, this solution is beyond the scope of science.
Date, Darwen & Lorentzos introduced 6NF. Equivalent form of 6NF is as follows:
"Relvar is in 6NF iff it consists of a simple key, plus at most one additional
attribute."
So, in fact, they declared that atomic structure is 6NF. In other words they gave
another name for the atomic structure. This name is 6NF. But they did not do only
important thing - they did not explain how to decompose data structures into atomic
structures.
The authors of "Anchor modeling" with their reliance on 6NF, show a serious lack of
understanding of important things in db theory.
In my papers I showed that db theory can not work as before but must be broken down
into two distinct theories:
(i) Simple Theory that only maintains and solves current states
(ii) General Theory that maintains and solves the "history", changes and other
things.
-- So I showed that the decomposition into atomic structures must be broken into two theory. In Simple Theory, it is Simple Form, and in General Theory it is the General Form. In my paper Simple Form, posted on May 15, 2006. on my website at www.dbdesign10.com , section 4 and 5, I presented Simple Form: ================================= 4.1 Definition. Relation schema R (K, A1, A2,...,An) is in Simple Form if R satisfies: R (K, A1, A2, ...,An) = R1 (K, A1) join R2 (K, A2), join ... join Rn (K, An) if and only if 1. Key K is simple 2. A1, A2,... , An are mutually independent. ================================== In my discussion with Derek (see thread "some information about anchor modeling") I explained, that K from Simple Form can be: 1. Industry-standard key. This type of keys is defined by international organizations. 2. Locally defined keys. This type of keys is defined by the local companies. For example telephone company uses its own system of keys in collecting money for their services. 3. Surrogate key. This is not about Cood's surrogate key. The surrogate key is stored only in db memory. In the first two cases keys are stored in the db memory and in the real world. Please note that "surrogate key" is acceptable for me. One form of surrogates I use in my solutions. But in the discussions on this user group I criticized Codd'sReceived on Tue Mar 31 2015 - 20:06:50 CEST
"invisible surrogates". After careful analysis, somewhere around 2005, I realized
that Codd's "invisible surrogate" is the wrong solution. So, when I criticize surrogate key, I think on Codd's "invisible surrogate". At this user group, people were discussing mainly the Codd's surrogate key. 4. Previous three cases belong to Simple db. When it comes to General db, then I use keys that represent identifiers of states. -- Based on the above cases, it is clear that 6NF does not exist, but there are the above mentioned four cases. These four cases allow decomposition of data structures in the atomic structures. -- In my paper "Some ideas about a new Data Model", Posted on my website at www.dbdesign10.com on September 17, 2005. , section 1.1, I wrote: "Besides Ack, every entity has an attribute which is the identifier of the entity or can provide identification of the entity. I already wrote that the following:" or can provide identification of the entity
"refers to a surrogate key, but more importantly, it does not have to be a surrogate
key, it can be anything that provides the identification of entities. So, again, the most important thing here is "it can be anything that provides the identification of entities." It is important to realize that in my data model, the identifier of entity can be represented with "K" that is specified in my Simple Form. =================================== Conclusion: The authors of "Anchor Modeling" made a big mistake. It is not possible
"decomposition" which is based on 6NF. In other words, does not exist
"decomposition", which can be done using 6NF.
There are only four, above mentioned decomposition. Three cases for Simple DB and one for General db theory. In my last post (26/03/2015) I showed that 6NF does not work with metadata, that means 6NF does not give the decomposition for General db. My General Form provides a decomposition for General db. My conclusion is that "Anchor Modeling" does not belong to science, because here I demonstrated that "Anchor Modeling" can not be based on scientific facts. =================================== What have the authors of "Anchor Modeling" plagiarized here? ************************************************************ They introduced "Historized and Static Attributes" and "Historized and Static Ties".
"Historized structures" belongs to my db General theory. "Static data structures"
belongs to my Simple db theory. However, I am the first(10 years ago), who introduced a new division in the theory of databases. General theory of databases is done using my theory of states. I introduced Simple form and Simple db that maintains current states. As far as I know, I'm the only one who did the decomposition for General db. I'm the only one who did the decomposition for Simple db. Vladimir Odrljin