|
Taxonomy and Hierarchical Data
Builder and Mapper / Triples tools are a set of
tree-control tools that enable tree viewing, creating, editing, and mapping
of data stored in conventional DBMS’ such as MS Access, SQL Server, and
Oracle. The two principal tools are Tree and Phrase. Tree is shown below for
a typical application.
Figure 1. Tree
Working
much like MS Window’s Explorer and similar tools, a left pane displays a tree
whose branches can be expanded and collapsed just as in Explorer, using the
+/- symbols to the left of the branch name. Just as in Explorer, branches can
be moved by drag-and-drop. Branches can be deleted, added, and renamed.
Again like Explorer, the right pane is a list view. The columns shown in the
list pane can be any field in the database record. The selection of fields
shown is similar to Explorer, via a right-click action on the column header.
Unlike Explorer, a Custom view can be defined by Customize popup selectable
at the right-click action on the column bar. With the popup, the fields to be
shown in the left to right order of the columns can be made quickly. Also,
the row height can be specified. Like Explorer, clicking on a column header
causes the list to sort in the order, click again
causes it to reverse order.
The tree is loaded by selecting a database upon which a popup is presented
that allows mapping of the name, description, and tree-parent fields. In case
there is a qualifier, e.g., a database code indicating a type of parent, that can also be specified. Upon mapping
completion, the tree is created. Any changes to the tree or database fields
are stored back into the database. The a field is added to the database or
the database is changed in some other way after loading in Tree, the tree can
be recreated using a Refresh (contents only change) or CreateTree
(structure change) button.
Phrase in a typical application is shown below. Phrase displays and allows
the entry and edit of all forms of mappings between trees. Phrase and Tree
can operate against the same database at the same time. The left and right
panes display the same tree although the branch expansions and collapses are
independent. The middle pane displays a mapping “type” tree. The mapping type
tree can be as simple (one node – “maps”) or complex as needed (e.g., a
domain’s predicates”.) Because of this, Phrase’s use can span from simple
mapping to elementary sentence or RDF triple creation to more complex
sentence and logical forms, e.g., triples where the object and predicates can
be triples and sophisticated predicate trees of logical or mereotopologic operators.
Why
These tools were created to increase the dimensionality of database user interfaces.
One of the powerful features of DBMS’ is the ability of users to view and
edit their data in very flexible ways interactively. No compilation, program
code, or batch processes are required. This enables a sort of end-user
computing, wherein end-users can rapidly create and iterate reports and data
analysis queries in what is relatively “real time” compared to a program
development cycle and with respect to the time from recognition of the need
to achievement of the result.
However, the dimensionality of the GUI is limited even when the
dimensionality of the data is high. High dimensional data occurs in all sorts
of applications:
- Part-of relationships in a supply database
- Type-of relationships in a catalog database
- Before-after in a personnel database
Relational
databases support high-dimensionality data with relationships, sometimes
many-to-many. However, the GUI for viewing and data entry out-of-the-box is
linear – that is rows vs. columns. There is no way to tractably create, view,
and edit multi-dimensional data and associated relationships with this type
of GUI.
Figure
2. Typical “datasheet” or query result GUI for DBMS’
Examples of
Navy processes in which these types of multi-dimensional GUIs have been
needed:
- Schema
mapping. Data schemas often have a
hierarchical structure, e.g., classes have attributes, messages have fields. Mapping schemas is an important
part of Community of Interest (COI) management under the Net Centric
Data Strategy guidance (DODI 8320.02G.) Mapping flat lists is
intractable for large and structurally different data schemas. Such
mapping is important for just about every COI and for many capabilities,
e.g., C2 and ISR.
- Enterprise
and Solutions Architectures.
As defined by DoD Architecture Framework (DoDAF) 2.0, hierarchies,
taxonomies, mappings, and complex multi-dimensional relationships are
required often to describe capabilities, business and operational processes,
services, systems, and implementations of solutions. For example, the
SV-5 relates System Functions (Activities performed by Systems) to
Operational Activities (Activities performed by Organizations).
Activities are almost always hierarchical, e.g., the Joint Chief’s of Staff Universal Joint Task List (UJTL),
the basis for operational training for all U.S. Forces. Another is the
development of a Systems Function hierarchy for the Performance
Specification of a new program, so that its function can be understood,
agreed-upon, and contracted for development. Phrase can then be used to
maintain different types of traceability to requirements and legacy
sources. There are many other examples for DoDAF and many of DTTS
features had their genesis in DoDAF and prior C4ISR architecture
work.
- Joint
Capabilities Integration and Development System (CJCSI 3170.01E). The JCIDS process requires development of DoDAF
views.
- Interoperability
and Supportability of Information Technology and National Security
Systems (CJCSI 6212.01E).
This process requires submission of DoDAF views so that interoperability
and supportability assessments can be made for programs in and of them selves and as parts of family or systems of
systems.
- Operation
of the Defense Acquisition System (DODI 5000.2). The DAS requires use of DoDAF views at milestone
reviews.
- Operational Situation Awareness and Data Fusion. Higher orders of data fusion associated with
situation awareness (e.g., object-object and object-event association)
require representation of high-dimensional data. Indeed, Phase was
initially developed under an ONR project for ontologically-supported
data fusion in which complex threat behaviors were entered in Phrase and
the reasoned-upon by a sophisticated multi-reasoning type application
built by the University of Buffalo’s Center for Multisource Information
Fusion (CMIF).
How
These tools are a set of applications that run under Microsoft’s .NET
Framework. The Microsoft .NET Framework is a software framework that includes
a large library of coded solutions to common programming problems and a
virtual machine that manages the execution of programs written specifically
for the framework. The class library is used by DTTS programmers, who combine
it with their own code to produce applications.
GitHub Development and Download Site: https://github.com/davem22101/OntologyTools_DataTree
and https://github.com/davem22101/OntologyTools_DataPhrase
|
