Re: The anatomy of plagiarism that was made by authors of "Anchor Modeling"
Date: Sat, 18 Apr 2015 00:12:01 -0700 (PDT)
Message-ID: <cd09e34e-7f02-481e-9b77-3cf67325c4df_at_googlegroups.com>
In my previous thread (from April 5th 2015), I showed the theory from my papers that provides an appropriate theoretical base for surrogates. Conditions that surrogates must meet also were shown. So, in my last thread, I presented a small theory related to surrogates. Authors of the "Anchor Modeling" did not do it. Conditions for surrogates E. Codd also did not present in RM / T. However, in my data model, I concentrated on the cases, which cover almost 100% of applications in the real industry. When it comes to surrogates, then I think that is enough to present the corresponding theory,
- In this post will be presented that the authors of the anchor modeling unclear
presented their data model. It is not clear what is their data model and what is not.
(i)
In their paper from 2009, which won first prize at the ER09, the authors of "Anchor Modeling", work simultaneously in two data models, ERM and RM. But in their paper is not done data mapping, at all. So it is not clear which data model they use, because it is impossible to use both models at the same time. In my Thread, "The original version", I wrote that the mapping between two data models is solved only in my data model. After my presentation of errors in "Anchor Modeling", on this user group, authors of "Anchor Modeling" published their second paper, in which they corrected mentioned errors.The authors of "Anchor Modeling" fixed these errors as follows: they were again plagiarized my results from my paper. This time they plagiarized my theory about "states" and even my technique. They plagiarized my identifier of states of relationships. We can notice that states of relationships is the most complex case with the states. The states are the most important part of my model. They published in Data & Knowledge Engineering journal, editor P. Chen.
(ii)
In their award-winning paper from 2009, the authors of "Anchor Modeling", wrote the
following: "In this paper, we propose a modeling technique for data warehousing,
called anchor modeling..." (see "Abstract", page 1).
(iii)
But on page 2, in this paper, they claim :"An Anchor model is a relational database
schema..."
(iv)
On page 13 they wrote "Anchor modeling is similar to Entity Relationship modeling.
(v)
On page 13 they wrote "Anchor modeling is similar to ULM".
(i)
On page 13, authors of "Anchor Modeling" wrote "Anchor modeling has several
similarities to the ORM".
(vii)
On page 2, they wrote :
"Although important, metadata is not discussed further since its use does not differ
from that of other modeling technique."
This is not true at all. "Metadata" work quite differently for General db, than in
conventional applications.
They keep the "History" and assist at identification. Metadata is much more used here
in General db than in ordinary databases. For example, in my databases for any data
that is entered in the database, I take without a user's knowledge, user
identification, time, date, id of the station from which it was entered. If there are
several procedures for data entry, then I enter the name of the procedure, with which
the data was entered.
In the process of creating identifiers, metadata can play a significant role. We can
see that neither Anchor modeling nor RM / T do not use metadata for identifiers. By
the way they use the keys, while I use identifiers instead of keys.
As I showed authors of anchor modeling did not work with "metadata" in their paper because they could not prove "decomposition" on the atomic structures and using of 6NF.
From these citations it seems that authors of "Anchor Modeling" have no clear idea of what exactly is their data model.
2. What my data model can model.
My data model can make a world. Of course here I do not mean some infinite world. It also does not mean the world in terms of religion, philosophy and so on. Here I am thinking of a small world that represents a whole and which can be used in industry, medicine etc.
The following two things related to the world are the most important:
(i) the world can maintain its past, present and future.
(ii) the main control part of the world is a collection of information, i.e. data
from the appropriate database.
Work with current states can be described using "Simple db " theory and the work with past states can be maintained using "General db" theory. Regarding current states and past states, on this user group there was a lot of discussion, and therefore I will not describe it again.
"History" of the future states can be realized as I described it in my paper: "Semantic Databases and Semantic Machines". In this paper I presented databases that actually work with sets of programs. These databases also maintained the state of the program and can work with the atomic structures of the program. Therefore, these databases can maintain parts of the program, that is parts of the corresponding processes.
I think this is a completely new topic that could be discussed more.
Vladimir Odrljin Received on Sat Apr 18 2015 - 09:12:01 CEST