Re: The original version

Data structures from Anchor Modeling are naively designed. Here is an example that confirms this.

Historized Tie in Anchor Modeling has the form: Htie(C1, C2, …, Cn, T). Then it is possible the following sequence of instances:
(C1, C2, …, Cn, T1)
(C1, C2, …, Cn, T2)

        ….
(C1, C2, …, Cn, Tk)
Now we can construct the algorithm which has the following rule: If i at Ti is an odd number, then the above relationship is “on”. Otherwise the relationship is “off”.

Now if a data entry person enters the following wrong instance: (C1, C2, … ,Cn, Ti) where T1<Ti<T2, then all the above instances which correspond to T2, T3, … , Tk become wrong.

There are many combinations in this simple example that show that Anchor Modeling is naive modeling. Just to mention that in the case of a crime, we can have an insider or we can have an insider and an outsider.
We can combine the above example with deleting of wrong data and as I have already shown - Anchor Modeling can not maintain “history” for deleted data.

For example, I remember the outages that happened in NYC, I think in summer of 2006. Many electrical lines were “off” many times and there are a lot of important companies, critical customers and government online agencies in NYC (the outages were “off” and “on” for about 10 days).

Vladimir Odrljin Received on Mon Dec 27 2010 - 01:44:46 CET