Semantics for Data Architects Fast Forward Video

A summary of the introduction to the ontology-derived Enterprise Data Model for Ontologists and Data Modelers.

Semantics for Data Architects fast forward is a summary of the FIB-DM technical introduction for Data Modelers and Ontologists.

The video is the second “Fast Forward” version of FIBO data model webinars and presentations. The equivalent to flipping through a deck of slides, the FF has the gist of the underlying education module. The viewer can decide to watch the full video and study the PowerPoint.

Watch Semantics for Managers, the non-technical fast forward for Finance and Business users first.

[advanced_iframe use_shortcode_attributes_only=”true” src=”https://www.youtube.com/embed/qKLtg1cA_Sg” width=”100%” height=”800″ id=”advanced_iframe” ]

You can read the presentation or download the PowerPoint here.

Transcript

Hello and welcome to this fast-forward of
Semantics for Data Architects the

Financial Industry Business Data Model
FIB-DM. Fast Forward is the

equivalent of flipping through a deck of
slides. It’s a summary of the two

full-lengths videos which together have more than an hour of content and

It’s meant to drive home the key
lessons learned and to give an overview

of what is in the decks and in the class.
Yeah, so, we make our case that FIBO is

the authoritative model of financial
industry concepts and definitions and

relations. And the main reason for that
is that FIBO is not a vendor model. It’s

created by an industry organization of
global financial institutions.

And it addresses the audience for this
class which can be a finance business

stakeholder. You can be a data or
application architect ontologist. And

about myself, I have 20 years industry
experience as the data architect and for

the last six or seven years as an
oncologist at leading financial

institution and service providers. And in
particular as an IBM Software Group

consultant for the Banking and
Financial Markets Data Warehouse, BFMDW,

I worked at 45 banks in North America
Europe and Asia implementing and

customizing the IBM industry model. So,
the core proposition is that there’s a

chasm between semantic and conventional
data management and basically this chasm

comes from the fact that FIBO in
ontology web language requires

specialized ontology specialized
databases which many midsize financial

institutions do not have in-house yet.
And even large institutions still must

support and design conventional
databases. So, and for these clients m

is the bridge across the chasm.
So, on the left hand side we have

FIBO deployed or triple stores in the
conventional word we have data models

and relational databases. The
configurable ontology to data model

transformation takes a FIBO RDF/OWL
import and generates a logical data

model. Then we go on about challenges
in leveraging the ontology for

relational design and the problems with
tooling support. Mainly that reverse

engineered ontology is to date; they are
not useful as a data model.

And the Financial Industry Business Data
Model is the FIBO in PowerDesigner and

other data modeling tools, consisting of
 entities definitions and notations

and axiom. It creates a common language and
design patterns for semantic and

relational databases. And our vision, our
goal is the semantic enterprise

architecture, where at the apex we have an
ontology the FIBO and we generate code

for our RDBMS, we generate a object code,
and in the future some messages and

processes. The way to do that is a
semantic model driven development, where

we start out with a FIBO, generate
conceptual models, logical models,

physical models, and deploy on the
infrastructure. Mid-sized financial

institutions adopt FIB-DM is a compatible
strategic enterprise model. Large

institutions can use CODT to transform
the in-house ontologies into data models

for downstream implementation.
We’ve covered transformation principles

and considerations for the derived model.
That must be practical; must be complete

fully documented; have diagrams; and it
must map back to the source ontology.

Now from FIBO to FIB-DM how does it work?
We cover the basic ETL process and its

using internally meta data sets for
ontology a generic entity relationship

and tools specific metadata. Now looking
at the two side-by-side we compare the

ontology graph to the conceptual data
model and we look at the

transformation of ontology elements to
data model objects. The conclusion is

that an ontology, a domain ontology
creates, generates a perfect conceptual

data model. We take a look at the
open source versus the commercial

version. And here is another perspective,
the import graph. At the top we have a

generic layer. In the FIBO, we have a
domain core of foundation, business

entities, finance business and commerce.
And we have extensions for securities,

derivatives and indicators. The upper
part, generic and domain core are free

open source the extensions are available
for licensing. And soon-to-come further

extension modules still in FIBO
development are loans, market data,

corporate actions, and collective
investment vehicles. Then we continue

part two of the education course. A demo
in PowerDesigner, looking at the package

structure, looking at package properties,

looking at extended attributes; that is
basically, the ontology the documentation,

that we import into the modeling tool. We
take a deeper dive at entity properties,

the naming convention,
and specialized tabs for harvesting the

ontology and annotations in the lineage. So,
ontology annotations are basically the

FIBO annotations converted into
PowerDesigner extended attributes, or

in Sparx tagged values, in ERWin
user-defined properties. the Lineage

provides a link to exactly identify the
URI of the otology item, what kind of

type it is and it captures class
restrictions

and equivalents as well. In the model we cover

multiple inheritance, entity attributes,
associations, and associative entities.

And we’ll take a look how ontology derived

models often come around as a kind of
a star schema design. And these Star schema designs

we can resolve these in the logical model. And relationships finally a merely linking

associative entities to the base
entities. Then we recap the chasm between

semantic and conventional data
management and FIB-DM as a bridge cross it.

Well thanks for watching this fast
forward of semantics for data architects

and you can go on the FIB-DM website to
download the data model, to watch the

full length underlying education videos,
and download the PowerPoint deck. Thanks

again, have a nice day