 |
Description  |
|
|
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention is directed generally to information management, and
more particularly, to systems for cataloging information managed by one or
more data processing nodes for use by non-data processing professionals
having access thereto. Still more particularly, the invention relates to a
system for providing catalog system product functionality within an
environment of fully user extendable meta-information objects.
2. Problem to be Solved
One of the challenges facing administrators of databases used by businesses
and other enterprises is to ensure that the information therein is fully
and efficiently available to and accessible by non-data processing
professionals. Knowledge workers are those persons within an enterprise
whose function is to make decisions, or recommendations, in response to
unanticipated events (e.g. changing business conditions). Those
individuals must have ready access to all relevant information available
to the enterprise, but may not know, or want to know, the unique naming
schemes, expression statements or semantic modeling required to retrieve
the information, particularly when that information extends over several
information management systems having differing storage and delivery
paradigms. Knowledge workers need to locate information on terms that are
understandable and intuitively apparent to them.
A requirement therefore exists for an information catalog system that
allows non-data processing professionals to define information objects on
their own terms, and to search, query, retrieve and otherwise manipulate
the "real world" information represented by those objects for
consideration and analysis. This presents a challenge to product
developers to produce a product whose objects are defined by the user base
but whose functionality remains fixed. End users must be free to generate
object types of their own creation but must be constrained in their
ability to assign product functionality thereto.
SUMMARY OF THE INVENTION
In accordance with the foregoing objectives, a computer implemented
information catalog database system is disclosed for cataloging
information stored in one or more data storage resources under the control
of one or more dam processing nodes. The catalog system includes a
cataloging service facility for performing one or more information
cataloging functions to organize and present a graphical view of the
information stored in the data storage resource. The information
cataloging functions are categorized into a plurality of defined function
categories. An object generation facility generates one or more meta-data
objects corresponding to units of information stored in the data storage
resource. The meta-data objects contain attributes defining
characteristics of the information units to which they correspond and the
mere-data objects are assigned to one or more of the function categories
to define the information cataloging functions which may be performed on
the meta-data objects. A user interface is provided for executing the
information cataloging functions on the meta-data objects in response to
user input.
BRIEF DESCRIPTION OF THE DRAWINGS
The objects, advantages and features of the present invention will be more
clearly understood by reference to the following detailed disclosure and
the accompanying drawing in which:
FIG. 1 is a system view of a data processing apparatus constructed in
accordance with the present invention;
FIG. 2 is a block diagram showing a set of function category classes
defined by the data processing apparatus of FIG. 1;
FIG. 3 is a block diagram showing the functionality assigned to the
function category classes of FIG. 2 and further showing a set of object
type subclasses generated by the data processing apparatus of FIG. 1;
FIG. 4 is a diagrammatic view of the structure of an information object
type class generated by the data processing apparatus of FIG. 1;
FIG. 5 is a diagrammatic view of an information object class generated by
the data processing apparatus of FIG. 1;
FIG. 6 is a diagrammatic view of an object instance generated by the data
processing apparatus of FIG. 1;
FIG. 7 illustrates a graphical user interface structure for creating an
information object type in accordance with the invention;
FIG. 8 illustrates a graphical user interface structure creating an object
instance in accordance with the invention;
FIG. 9 illustrates the structure of a tag language file for generating an
information object type subclass in accordance with the invention;
FIG. 10 illustrates a graphical user interface structure for selectively
executing one or more cataloging services in response to user input;
FIG. 11 illustrates a graphical user interface structure for executing a
database catalog search in response to user input;
FIG. 12 illustrates a graphical user interface structure providing a list
of object instances resulting from a database catalog search;
FIG. 13 illustrates a graphical user interface structure for navigating a
selected set of object instances in the database catalog; and
FIG. 14 illustrates a graphical user interface structure showing an
information object type created by the data processing apparatus of FIG. 1
.
DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
SYSTEM ENVIRONMENT
Referring now to FIG. 1, a data processing system 2 includes one or more
information processing nodes 4 which may function as independent data
processing devices or may be part of a network of shared data processing
resources. The data processing system 2 will be populated in many cases
with an assortment of digital computing devices, data storage libraries
and other resources, including but not limited to, personal computers,
work stations, mid-range computers and computer mainframe systems, all of
which are preferably configured to share information with other resources
in the data processing system.
In the system of FIG. 1, the data processing node 4 is a desktop IBM
personal computer 6 having a microprocessor controller, operational
program memory and one or more disk drives for permanent program storage.
The data processing node 4 further includes a video monitor 8, a keyboard
10 and a cursor control device such as a mouse (not shown) for controlling
a visual cursor representation displayed on the video monitor 8. As is
known, the cursor can be used for making selections among graphical
objects displayed on the video monitor 8, which objects represent
functions provided by programs installed and running on the personal
computer 6. An operating system, such as OS/2 2.1 from IBM, is installed
on the personal computer 6 to manage system resources and assist in
program execution.
The data processing node 4 is connected for information exchange with one
or more data storage resources, each of which could be a data processing
node running under its own operating system. For example, one node could
be a mainframe computer system 12 running under an MVS operating system
and managing a database such as IBM's DB2 catalog product. Another node
could be a personal computer 14 running under IBM's OS/2 2.1 operating
system and also managing a database such as IBM's DB2 catalog product.
Still another node could be a mainframe computer system 16 running under
the VMS operating system and managing another database product. Still
another node could be a work station 18 running under the Unix operating
system and managing another database product. Other nodes 20, 22 and 24
could be added as desired and could represent computer systems managing
additional databases. They could also represent computers running other
database catalog systems of the type described herein or could further
represent other sources of meta information.
As used herein, "meta" information is any user defined information adopted
as a convenience for describing data processing entities such as records,
files and other objects contained in databases or other data processing
resources. Meta information is the user's own description of object
information which can be used to search, query, retrieve and otherwise
manipulate enterprise information independently of the unique naming
schemes, expression statements or semantic modeling imposed by the data
processing resources that contain and manage the information.
In the system of FIG. 1, the data processing node 4 is configured for
information exchange with the data processing nodes 12-24. The data
processing node 4 is programmed to provide a database catalog system 26
for storing and manipulating meta information contained in a metadata
store 28. The metadata store 28 is a local data storage resource that
utilizes a memory device such as a disk drive to store meta information
objects corresponding to information managed by one or more of the data
processing nodes 12-24. The database catalog system 26 manages the
metadata store as a relational database and provides sort-query-logic
(SQL) support for accessing the meta data objects therein. The database
catalog system 26 retrieves information from the data processing nodes
12-14 via its import facility 30, its tag language file storage facility
32, and various external extract facilities 34.
The facilities 30-34 represent functional services and resources available
to the database catalog system 26, and operate in a manner now to be
described. When a user of the database catalog system 26 desires to create
meta information representations of information managed by one of the data
processing nodes 12-20 and 24, an extract facility 34, which would be
normally provided by a program running under each data processing node, is
activated to retrieve such identification information from the data
processing node as may be required to selectively retrieve units of
information managed by that data processing node. The extract facility
preferably includes a tag language facility that formats the
identification information into one or more tag language statements for
use by the database catalog system 26. The tag language statements are
stored as a file in the tag language storage facility 32, which is a
resource provided by the database catalog system 26. The import facility
30, which is a service provided by the database catalog system 26,
interprets the tag language statements generated by the tag language
facility 32 to generate meta information objects for storage in the
metadata store 28.
When a user of the database system 26 desires to obtain information from
the data processing node 22, a modified procedure is used. The data
processing node 22 is a remote database catalog system which already
contains meta information and is configured with its own local export
facility 36. The export facility 36 converts meta information objects
managed by the remote database catalog system into tag language statements
that are placed in a file. Those tag language statements can then be
routed around the tag language facility 32 and utilized directly by the
import facility 30 for conversion into local meta information objects.
OPERATIONAL OVERVIEW
Referring now to FIG. 2, the database catalog system 26 allows knowledge
workers to define meta information object types and object instances
thereof corresponding to information managed by one or more data storage
resources under the control of one or more data processing nodes, such as
the nodes 12-24. In large measure, the user-defined object types are
customizable and extendable by knowledge workers or data processing
administrators thereof. The users determine what information should be
captured, what it is to be called, and how it is organized. The database
catalog 26 has no foreknowledge of the object types users will generate.
However, the database catalog 26 must support the user-generated object
type with a defined set of database catalog functional services that are
appropriate to the object types generated by users.
To satisfy that objective, the database catalog system 26 defines a
plurality of functional categories within which the user-defined object
types may be generated. Each functional category represents a
categorization and/or subclassing of a super class of functional services
provided by the database catalog system 26. The functional categories to
which the object types are assigned become part of the object type
definition and limit the functions available to each object type. An
object type is thus a data structure that may be thought of as a subclass
object that encapsulates the functions inherited from the category class
to which the object type belongs, together with one or more property
attributes corresponding to information that the knowledge worker wishes
to catalog. The object types can be populated with object instances that
are generated by assigning values to the property attributes to create
meta information objects that uniquely identify units of information to be
cataloged by the knowledge worker.
In the database catalog system 26, the functional categories to which the
object types may be assigned include, by way of example only, a "Grouping"
category 40, an "Elemental" category 42, a "Contact" category 44, a
"Dictionary" category 46, a "Support" category 48, and a "Program"
category 50. All of the user-defined object types of the database catalog
system 26 are placed in one of these six categories. Each category
represents a distinct set of product functionality. The functional
categories 40-50 may be further divided into two groups. A first group
represented by categories 40, 42 and 50 contains "Information" objects
corresponding to information from one or more data storage resources to be
cataloged. A second group represented by categories 44, 46 and 48 contains
"Support" objects corresponding to information that supports the
cataloging of information defined by the Information objects. Examples of
object types that may be assigned to the object categories 40-50 are
illustrated in FIG. 2. These object types may be predefined within the
database catalog system 26 and provided as an object type "starter set"
for user convenience. It will be understood, however, that many other
object types may be defined and users may well choose to delete the object
type starter set in favor of a completely user-customized set of object
types.
For purposes of discussion, however, a database catalog system is assumed
to be populated with an object type starter set only, in which case the
Grouping category 40 will be understood to contain a "Business Group"
object type and a "Tables" object type. These object types represent
information groupings containing information units or elements as members
thereof. The purpose of the Business Group object type is to provide a
generic container for organizing other grouping and elemental category
object types such as tables, columns, queries, reports and images. The
purpose of the Table object type is to describe the relevant properties of
an SQL relational table or a client/server file.
In the example of FIG. 2, the Elemental category 42 contains "Column,"
"Data Analysis," "Completed Report" and "Image" object types. These object
types are the individual members that could make up a Grouping object
type. For example, the Column object type would complete the table
definition established by a Table object type in which it is contained.
The Data Analysis object type is used to represent some previously
formulated analysis of data which a knowledge worker can request an
informational application to perform. The analysis could be in the form of
a query, a chart, a report, a spreadsheet, or an SQL request. The
Completed Report object type represents fully processed information that
is ready for presentation to a knowledge worker. The purpose of the Image
object type is to represent image information such as that contained in
blueprints, maps, schematics and other graphical representations.
In FIG. 2, the Contact category 44 contains a "Contact" object type whose
purpose is to identity persons that can provide support and assistance for
one or more Grouping or Elemental object types. The Dictionary category 46
contains a "Glossary" object type whose purpose is to provide
clarification of terms used by other object types. The Support category 48
contains a "News" object type whose purpose is to provide a mechanism for
knowledge workers to advise other users of the current state of the
database catalog. The Program category 50 contains a "Program" object type
that defines application programs capable of processing particular
Grouping or Elemental object types. Advantageously, as described in more
detail below, various relationships can be defined between object
instances from certain related function categories for cross-referencing
information provided by the object instances.
FUNCTION CATEGORIZATION SUBCLASSING
Referring now to FIG. 3, an object oriented view of the database catalog
system 26 is provided to illustrate some of the product functions that may
be assigned to each of the functional categories 40-50. FIG. 3 shows a
hierarchy of classes and subclasses encapsulating one or more product
functions or "methods" together with "properties" defining characteristics
of the information represented by the object types.
The database catalog system 26 contains a catalog service facility 60 for
performing a plurality of information cataloging functions to organize and
present a graphical view of information stored in one or more data storage
resources. Some of these information cataloging functions are categorized
in a mutually exclusive fashion by the manner in which they are assigned
to the function categories defined above. In addition, there is a shared
set of basic product functions represented by a database catalog function
superclass 62. The superclass 62 contains a plurality of functional
methods and a set of attributes or "properties" that are common to all
object types that inherit the methods and property attributes of the
superclass. The functional methods of the superclass 62 are the "Create,"
"Get," "Update," "Delete," "ListObjectTypes," "SearchAll," and "Search"
methods. The defined superclass properties are a user provided Name, an
Object Type Identifier, an Object Instance Identifier, a Last Modification
Time and Date Identifier, and a Last Modification User Identifier.
The Create method is a routine that allows knowledge workers to create
object types and object instances. The Get method allows knowledge workers
to retrieve information about previously created object types and object
instances. The Update method allows knowledge workers to modify previously
created object types and object instances. The Delete method allows
knowledge workers to delete previously created object types and object
instances. The ListObjectTypes service allows knowledge workers to
generate a list of all previously created object types. The SearchAll
method provides a fuzzy search capability on any of the properties of
object instances within a given object type. The Search method allows
knowledge workers to search for object instances containing a specified
property attribute across all object types.
The database catalog superclass 62 is subclassed into a fixed set of
abstract object subclasses which are the functional categories 40-50
previously described, together with an internal subclass 64 that is
subclassed into the Grouping and Elemental category subclasses 40 and 42.
As can be seen in FIG. 3, the Contact, Dictionary and Support category
subclasses 44, 46 and 48 do not provide additional methods beyond those
contained in the superclass 62. The Contact, Glossary and News Object
types 66, 68 and 70, which are created as subclasses of the Contact,
Dictionary and Support category subclasses, inherit the methods and
property attributes of the superclass. Moreover, knowledge workers may
define additional property attributes for these and other object types
created as subclasses of the Contact, Dictionary and Support category
subclasses. These category subclasses are also extendable in that they
permit knowledge workers to generate additional object type subclasses in
addition to the starter set provided with the database catalog system 26.
The Program category subclass 50 does not contain any additional functional
methods but does define additional property attributes that are inherited
by the Program object type subclass 72. These additional property
attributes include the Program Names of available application programs,
the Object Types this Program Handles, the Program Startup Commands, and
the Parameters required to invoke the programs using data resource
information represented by the Grouping and Elemental object instances.
Knowledge workers may also define additional property attributes for the
Program object type 72. Unlike other function categories, however,
knowledge workers may not create additional object types in the Program
category, but may generate multiple instances of the Program object type
72.
The internal subclass 64 adds the "Relation," "ListContacts" and
"ListPrograms" methods to the methods inherited by the Grouping and
Elemental subclasses 40 and 42 and the object type subclasses thereof. The
Relation method is used to establish a "contains" (i.e. parent-child)
relationship between object instances that inherit the characteristics of
the internal subclass 64, with the restriction that object instances that
are further subclassed as Elemental object instances are always the target
(e.g. child) of a contains-relationship established by the Relation
method. The Relation method also permits a relationship to be established
between object instances that inherit the characteristics of the internal
subclass 64 with object instances that inherit the characteristics of the
Contact category subclass 40. The ListContacts method reports on the
status of this relationship by identifying the Contact object instances
relating to a Grouping or Elemental object instance. Relatedly, one of the
property attributes of a Program object instance is the object type the
program handles. The ListPrograms and Open methods are provided to report
on the relationship between Program object instances and the Grouping and
Elemental object instances they specify.
The Grouping category subclass 40 adds the "ListAnchors," "Navigate" and
"WhereUsed" methods to the Grouping object type subclasses 74 and 76 that
inherit the characteristics of the Grouping category subclass. The
ListAnchors method identifies all those Grouping object instances that are
not contained in any other Grouping object type subclass (i.e., they are
at the root of some classification scheme). The Navigate method identifies
all object type instances that are directly contained within a specified
object instance of a Grouping object type subclass. If any of the returned
object type instances contain additional object type instances they are
flagged to indicate that the Navigate method could be applied to those
object instances as well. The WhereUsed method identifies all object
instances that directly contain a selected object instance.
The Elemental category subclass 42 adds the WhereUsed method to the
Elemental object type subclasses 78, 80, 82 and 84 that inherit the
characteristics of the Elemental category subclass 42. This is the same
WhereUsed method provided for the Grouping category subclass 40, except
that Elemental object instances are barred from using the Relation method
to relate to another object instance that has inherited characteristics
from the Grouping or Elemental category subclasses.
DATABASE CATALOG IMPLEMENTATION
The provision and categorization of cataloging functions by the database
catalog system 26 may be accomplished in several ways by persons skilled
in the art based on the disclosure herein. The system 26 is preferably
implemented by a general purpose computer, such as the data processing
node 4 of FIG. 1, which executes a sequence of processing commands
provided by a database catalog computer program. The program could be
written in a variety of computer languages including the well known "C"
programming language. In a preferred embodiment, the cataloging functions
of the database catalog system 26 are implemented as a series of C
language functions. Those functions preferably provide sort-query-logic
(SQL) database access capability. As described in more detail below, the
object types and object instances are generated as data structures having
a plurality of defined fields corresponding to contiguous memory locations
that store the information which defines the object types and the
instances thereof. One of the data structure fields identifies the
functional category of which the object type or object instance is a
member. When a call is made to one of the cataloging functions that
operate on object types and instances, the procedure reads the category
identifier of the object type or instance data structure. If the category
identifier is recognized, the function is executed. If not, an error
message is displayed advising that the requested procedure is not
available for the object type or object instance identified. In this way,
the meta objects of the database catalog system 26 encapsulate the
functions and properties which they inherit from the function category
classes to which they belong. Such encapsulation could be similarly
achieved using object oriented programming tools such as the C++
programming language.
META OBJECT DATA STRUCTURES
Referring now to FIG. 4, an exemplary structure of a meta data information
object 90 of the present invention is shown. The information object 90 is
a C language data structure having three contiguous parts, a header area
92, a definition area 94 and an object area 96. As stated above, the areas
92, 94 and 96 contain fields corresponding to contiguous memory locations
for storing information about the object 90 represented by the data
structure.
The header area 92 contains information about the nature and size of the
data structure and includes a structure identifier 98, a definition area
length identifier 100 indicating the size of the definition area, an
object area length identifier 102 indicating the size of the object area,
an object area entry count identifier 104 indicating the number of
property values contained in the object area, a category descriptor 106
specifying which of the functional categories contains the object 90 as a
subclass member thereof (there is only one), an object type identifier 108
and a reserved area 110.
The definition area 94 contains information about the property attributes
of the object 90 and includes, for each property attribute, a property
name identifier 112 containing the name of the property, a data type
identifier 114 indicating the data type of the property, a length value
116 indicating the size of the property, a physical property name 118
indicating a physical name of the property, a value flag 120 indicating
whether a value for the property is system-provided, user-supplied or
optional, a universal unique identifier sequence number 122 indicating
that the property is a member of a sequence of a set of universal unique
identifiers that can be used for identifying the object 90 in relation to
other objects in the database cataloging system 26 and in relation to
other database cataloging systems and meta data stores, a case sensitivity
flag 124 used for the Search function to distinguish between upper and
lower case property values, a fuzzy search flag 126 used for the Search
function to specify fuzzy or nonfuzzy searching, and a reserved area 128.
The object area 96 is used to contain values for the properties defined in
the definition area 94. It contains plural fields, such as the value
fields 130, 132 and 134, for identifying the property values of the object
90.
Referring now to FIG. 5, an object type 140 of the Grouping category
subclass 40 is shown. The object type 140 contains a header area 142 and a
definition area 144, but no object area. Object types do not have object
areas because they do not contain property values. Only object instances
have assigned property values. The properties defined for the object type
140, which may be a Business Group object type, include system generated
properties 146, 148, 150 and 152 representing the four global properties
of the superclass 62 described above. The properties of the object type
140 further include a set of required properties provided by the knowledge
worker, including the Name property 154 and the Unique Universal
Identifiers 156, 158, 160, 162 and 164. Additional properties that may be
optionally assigned, deleted, extended or otherwise manipulated include
the Actions property 166 that specifies application programs capable of
using the object type 140, an Extract Source Type Token 168 identifying
the extractor program/process that created the tag language file used to
populate the object type 140 with object instances, a Short Description
170 and a Long Description 172 providing additional information about the
object instances populating the object type 140. Many other properties
could also be provided as required to support the cataloging of meta
information for use by knowledge workers.
Referring now to FIG. 6, an object instance 180 of a Business Group object
type is shown. The object instance 180 includes a header area 182, a
definition area 184 and an object area 186. It will be appreciated that
the fields of the object area 186 sequentially correspond to the
properties identified in the definition area 184. Thus, the object area
field 188 contains a six character Object Identifier value. The adjacent
Instance Identifier field 190 is left blank by the system for this object
instance. The Name field 192 contains a twenty-one character name. The
adjacent Last Changed Date and Time and Last Changed By fields 194 and 196
are left blank by the system for this object instance. The Source field
198 contains a three character value indicating the extract source type.
The Short Description and Long Description fields 200 and 202 are filled
with multi-character values that provide additional meta information
regarding this quality group instance.
USER INTERFACES
The database catalog system 26 can be configured to permit object type and
object instance creation, deletion, editing and related manipulation in
several ways. FIGS. 7 and 8 illustrate the operation of a graphical user
interface system that allows knowledge workers to perform such functions
using the video display 8, the keyboard 10 and the mouse input device
previously described. FIG. 7 illustrates a graphical user interface panel
210 for creating an object type. The category subclass, various names and
tags, together with the properties of this object type may all be
specified in an interactive graphical input session. FIG. 8 illustrates a
graphical user interface panel 220 for creating an object instance. The
object type may be specified, together with the values of the properties
defined by the object type, in an interactive graphical input session.
Other graphical user interface panels for manipulating object types and
object instances may also generated by the database catalog system 26 to
provide functionality similar to that of the panels of FIGS. 7 and 8.
FIG. 9 illustrates an alternative batch input method using a series of tag
language statements 230 for defining an object type. Object instances may
be defined in similar fashion. A third alternative input method is to
utilize an application programming interface (API) allowing access to the
database cataloging system via an external application using C language
function calls.
Once the metaadata store 28 is generated by creating one or more object
types and object instances in accordance with the methods described above,
database cataloging system 26 allows knowledge workers to access the meta
information contained therein within the context of an interactive
graphical user interface environment. Referring now to FIGS. 10-14, a
series of exemplary graphical user interface panels intended for use by
knowledge workers to execute various database cataloging and display
functions are shown. FIG. 10 illustrates a main work area window 240 in
which one or more database cataloging functions, such as the Search
function, may be represented as user activatable icons. Additional icons
may be provided to indicate stored searches which may be viewed by
selecting the appropriate icons. FIG. 11 illustrates a search window 250
for implementing the Search procedure following selection of a Search icon
in FIG. 10. FIG. 12 illustrates a window 260 containing the results of a
previously executed search. FIG. 13 illustrates a window 270 containing
the results of the Navigate method using a platform tree-view paradigm as
the preferred way to graphically display the contains-relationships
between related object instances. FIG. 14 illustrates a window 280 for
graphically displaying a detailed description of an object instance. It
will be appreciated that the window 280 provides a mechanism for viewing
all of the meta information provided by the object instance, and it is
this window whi | | |
