The basic business analysis for IBIS was
conducted over a period of two years and was developed from the
analysis of several, disjoint, sources including retail, service and
manufacturing oriented businesses and medical clinics, as well as
from conversations with business people. The business model that
emerged was an I/O model of business - what comes into and goes out
of a business.
Our analysts made a list of each activity of a
typical business, along with the resources used by that activity,
and the information needed to manage that aspect of the business,
including activities both internal and external to the business.
The business model itself is based on a
General Systems model. To use the purchasing process as an example,
our analysis for the purchasing process followed it from the issue
of a requisition by a department, through issue of a purchase order,
receipt of goods, and delivery of the goods to the department. The
various actors, actions, goals, and information flows were
identified, and then a model was built from them. This then resulted
in a database structure and appropriate screen designs to support
it.
Comparison of our model to those developed by
the National Retail Federation, and several universities has
revealed our model to be more comprehensive, as the other models
focus on specific industries, or exclusively on e-commerce. We have
also investigated business and enterprise ontology projects, such as
the one at Stanford, to see if we've missed anything.
Our analysis of the way businesses actually
operate was thorough and comprehensive. What emerged at every step
was a model that was flexible enough to handle all of the various
permutations that different types of businesses employed to
accomplish a stated task. For example, in discussing till drawer
assignments, we identified eighteen standard patterns of assignment,
twenty-one methods of close out, and about a dozen types of
emergency situations and approximately five ways to handle each one.
In response we constructed an elegant solution (the fewest steps and
least confusion) for IBIS to handle all of these permutations with
no changes in the way these businesses currently managed their
tills.
Then we designed the screen, using not just
the Apple, Microsoft and DOD Human Interface guidelines, but also
those design elements used by the learning disabled (color blind,
dyslexic, etc.) to be sure that as far as possible, the screen would
be clear and the information flow obvious to them with minimal
training or alteration of the screen.
Our analysts looked at such things as
evidentiary rules for auditing for theft and/or fraud, and the needs
of the Labor Commission, FDA and other agencies that would use this
for forensic reasons, as well as numerous industry specific
situations, and made sure that that IBIS could be used to handle
these with little or no customization.
Then we drew out every permutation of each
step and tried to find out who might need that information. Once we
had a comprehensive list of every information point anyone might
need, we made sure we gathered, in the finest detail, everything any
business might need. Finally, we returned to the regulations, such
as tax codes, USDA Perishables Handling, etc. to insure that IBIS's
design included all that these regulations and standards specified.
Our design team operated under the assumption that if IBIS handled
everything required by the Federal regulations and those of the
State of California it would handle any other state as well, since
the State of California is known as a regulatory nightmare for most
businesses.
1
In philosophy, ontology is the study of the kinds of
things that exist. Ontologies are often said, colorfully, to "carve
the world at its joints."
In Artificial Intelligence
contexts, the term has largely come to mean one of two related
things:
A representation vocabulary,
typically specialized to some domain or subject matter. More
precisely, it is not the vocabulary as such that qualifies as an
ontology, but the conceptualizations that the terms in the
vocabulary are intended to capture. For example, the ontology
doesn't change by translating the terms from pseudo-English to
pseudo-French.·
Occasionally, a body of knowledge
describing some domain, typically a common sense knowledge
domain, using such a representation vocabulary. For example, CYC
often refers to its knowledge representation of some area of
knowledge as its ontology.
