Data processing: software development – installation – and managem – Software program development tool – Code generation
Reexamination Certificate
1993-08-27
2002-10-22
Morse, Gregory (Department: 2124)
Data processing: software development, installation, and managem
Software program development tool
Code generation
C707S793000
Reexamination Certificate
active
06470490
ABSTRACT:
FIELD OF THE INVENTION
The present invention relates to the field of information representation and retrieval within a general-purpose digital computer. Information of all simple (one-dimensional) types is represented as points which are arrayed along dimensions (vectors), and compound information types are represented as the intersection of two or more dimensions in a multidimensional data space. A context, comprising one or more points may be established and maintained locally or globally, for use in evaluation of values within the data space. An evaluator function is used to return values which are bound to points in this data space, and to invoke conventional data processing functions which interact with the data space. The evaluator resolves values by selectively incorporating points from the context on an as-needed basis. Using the contextual representation and retrieval method, the processes and structures of conventional computing, such as variables, arrays, structures, lists, objects, and the like may be modeled or simulated.
BACKGROUND OF THE INVENTION
The representation of information within the memory of a general-purpose digital computer, and the organization and retrieval of that information is one of the classical problems of computing. The mere representation of alphanumeric characters which has become so commonplace was a major advance which required a complex mapping of (at first) seven bit values to corresponding characters and meta-characters in a standard, accepted order. The one-to-one mapping of the values [0,1,2 . . . 126] to the set of characters [NUL, SOH, STX . . . ~] was not a simple task to standardize. It was not until 1965 that such a mapping became the ASCII standard.
The construction of data files using computers began with the sequential file, in which a series of records of a known and fixed length were concatenated together, and could be retrieved by the computation of an “offset” into the file, to arrive at the desired record. Thus, to arrive at the nth record in a file consisting of 80 byte records, one merely discarded the first (n×80) bytes read from the file, and retrieved the 80 bytes immediately following.
Later developments in data retrieval and storage technologies included the indexed sequential file which permitted variable length records without the need for sequential search in order to retrieve a desired record, and most recently, relational databases which use a common “key” field to relate multiple “tables” of information.
When coupled with a high level language capable of manipulating information and indices, the relational database can be a powerful tool for managing information. Each table is usually defined as a two-dimensional array of information, having columns which correspond to fields having a defined type, such as alpha, date, integer, real, and so on, while the rows correspond to individual records within the database. Relations are formed by the matching of values within designated fields of different tables. For instance, a typical business database may have a table for storing information about “people” comprising fields such as: Customer Number, First Name, Middle Initial, Last Name, Street Address, City, State, Zip Code, Telephone Number, and Social Security Number. Another table may list items purchased, and may include fields such as: Item Number, Number Purchased, Date Purchased, Item Price, Discount Applied, and Purchaser Customer Number. By knowing the Customer Number of a given customer (perhaps retrieved after a search for the customer's name in the “people” table, a listing of every item purchased by the customer may be retrieved from the “items purchased” table. Further relations are then possible in such an extended system to determine, for instance, the identity of the manufacturer of a particular item, the count of that item in current stock, and other information necessary to operation of a business.
Packaged together with tools for report formating, data manipulation, and user interface construction, relational databases are sometimes referred to as “Fourth Generation Languages” or simply “4GLs”. While such 4GLs have proven application in business computing, they are inadequate for certain “real world” tasks, and are cumbersome for others.
The present invention comprises a method for database retrieval which may be termed “super-relational”. It provides for an arbitrarily large number of simple information storage structures which may be combined in large numbers of ways, and economically stored and accessed using convention general-purpose digital computers (as opposed to special-purpose database systems such as the GScan or Teradata computer systems.)
BRIEF DESCRIPTION OF THE INVENTION
In the method of the present invention, all informational entities are treated as points along dimensional lines. For instance, every integer number is treated as a point along an “integer” dimension, every real number as a point along a “real” dimension, and every breed of dog as a point along a “breed_of_dog” dimension. In fact, dimensions can exist for any concept or item which requires representation in a data processing system. All that is required is that every point along a given dimension possess the same quality of interest as all other points along that dimension (consider such dimensions as “red_things” and “fruits”).
A point in the method of the present invention is of the notational form:
dimension: label
where dimension is the name of the dimension (e.g., integer, real, or fruits), and label specifies a particular point along the named dimension (e.g., “3”, “3.14159”, or “apple”).
A binding in the method of the present invention is an association of a value point &ugr; with the intersection I of two or more other points p within the multidimensional space, so long as no two points are contained along the same dimension. A binding is represented by the notation:
[
p
1
, p
2
, . . . p
n
]=&ugr;
where the intersection I consists of the points p
1
, p
2
, p
n
enclosed in brackets being bound to the value point &ugr;, and defining the intersection at value=&ugr;(i.e., &ugr; along the “value” dimension).
It is thus possible to reference a value &ugr; by specifying a set of points p
1
, p
2
, . . . p
n
. A complete intersection is one in which all p
n
points are specified, while an incomplete intersection is one in which one or more points are missing.
The context set C in the method of the present invention is a set of points c such that no two points belong to the same dimension. A context is represented using the following notation:
[
p
1
, p
2
, . . . p
n
|c
1
, c
2
, . . . c
n
]
which means that the method will evaluate the intersection p
1
, p
2
, . . . p
n
within the context of points c
1
, c
2, . . . c
n
. If the intersection specified in p
1
, p
2
, . . . p
n
is complete, then no further operation need occur. However, if that intersection is incomplete, then one or more points taken from the context C may be used to complete the intersection. For instance, the binding:
[
a, b, c
]=value
cannot be referenced as [a, b] because the intersection is incomplete. Given a context of
[
a, b |c, d, e]
however, the intersection may be evaluated by using c from the context.
At the core of the method of the present invention is the evaluator E, which is a process which, given an intersection I, determines the value &ugr;, provided that I is a complete intersection, or is provided with a context C. That is,
E
(
I
)=&ugr; or
E
(
I,C
)=&ugr;
The context C is preferably maintained externally to the evaluator, and points from the context are selected for use in evaluations on an as-needed basis. The evaluator must also be able to resolve ambiguities, that is, it must be able to choose between two or more possible (and complete) intersections in order to return a single value. Several approaches, such as simple point weighting are possible, and have been proven to be effective in t
Chavis John Q.
Husick, Esq. Lawrence A.
Lipton, Weinberger & Husick
Morse Gregory
LandOfFree
Contextual data representation and retrieval method does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Contextual data representation and retrieval method, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Contextual data representation and retrieval method will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2989354