Dec 172017
 
Weighted Data Retrieval for Dbase III/III+.
File INTEL.ZIP from The Programmer’s Corner in
Category Dbase Source Code
Weighted Data Retrieval for Dbase III/III+.
File Name File Size Zip Size Zip Type
IT.DOC 19383 6528 deflated
IT.EXE 183168 73193 deflated

Download File INTEL.ZIP Here

Contents of the IT.DOC file






INTELLI-TRIEVE (tm)

Copyright 1987, Richard Ray Viets, All Rights Reserved

Intelli-Trieve is a weighted retrieval utility for users of dBASE
III, dBASE III+ (trademarks of Ashton-Tate), and compatible
database managers. The weighted retrieval capability provided by
Intelli-Trieve is a powerful and versatile tool that performs a
very complex task easily - without programming.

Weighted retrieval has been available for years to users of
powerful mainframe systems like IBM's CICS. Intelli-Trieve now
brings this powerful concept to microcomputer users.

Intelli-Trieve performs weighted retrieval by comparing each
record in a database to user-defined selection criteria, and
producing a copy of the database with the records arranged in
descending order of 'acceptability'.

The result is an entirely new 'view' of your data, one that cannot
be attained through standard indexing, filtering or sorting
operations. Intelli-Trieve allows you to extract new and valuable
information from your data in ways that were not considered in the
original database design - without altering the structure of your
databases.

Intelli-Trieve's simple, menu-driven process allows you to choose
a database, indicate which fields are to be included in the
weighted retrieval calculations, and how each field is to be
judged and weighted.

Intelli-Trieve then reproduces the database, ordered by the new
selection criteria, after which you can apply your DBMS's browsing
and reporting capabilities to view your new information.

The goal of Intelli-Trieve is to provide an easy way to extract
new and valuable information from existing data. You can use it
immediately, and in minutes see an increase in the value of your
data.

Intelli-Trieve doesn't try to do those things your DBMS already
does so well, and which you've invested time and effort to master.
It provides added value instead of reinvented wheels.

What are some of the ways you can apply the power of Intelli-
Trieve? Weighted retrieval is well suited to aiding you in
selecting securities investments based on your definition of the
relative importance of factors such as volatility, price, price-
to-earnings ratios, volume -all considered and weighed
simultaneously.

Another use is to perform any sort of complex and detailed
preference matching, from the traditional dating service to


-1-










job/candidate matching, business site selection or student
admissions.

Sales people can use Intelli-Trieve to develop efficient leads
based on demographic and geographic criteria they set to match the
product and its market.

Executives responsible for purchasing decisions can rely on
Intelli-Trieve to simplify the complex task of weighing the
relative merits of hundreds of competing products, based on dozens
of independently considered criteria.

With Intelli-Trieve existing databases, designed for yesterday's
information needs, can be made to reveal new information, based on
the demands of today.

A WEIGHTED RETRIEVAL EXAMPLE

A simple example of how weighted retrieval might be used in
solving a marketing problem will demonstrate Intelli-Trieve's
power and ease of use. The object in the example is to extract
qualified sales leads from a database consisting of thousands of
records, each of which contains varied - and often incomplete-
demographic data about one person.

Suppose the product we wish to sell is a fast, expensive sports
car which only a very small portion of the people represented in
our database can afford, or would even want. In order to maximize
our efforts we will want to contact the most likely candidates
first, in the order of their sales potential, and avoid wasting
our time on those unlikely to buy our product.

We might know, for example, that men are more likely to indulge
themselves with a sports car; especially if they are single. We
also know that our chances of making a sale increase if our
prospect earns $40,000 a year or more, and we have an even greater
chance of making a sale to someone making over $60,000. Suppose we
also think that prospects within three months of their fortieth
birthday may be especially susceptible because of 'midlife
crisis'. And finally, we know we will have greater success among
those whose last automobile purchase was at least two years ago,
especially if the purchase price of the old car was over $8,000.
If that purchase was not financed we stand an even better chance
of making the sale.

The standard methods of filtering or indexing might be able to
extract from the data those who met ALL our criteria, but precious
few sales leads would result. Worse still, many excellent leads
would be lost altogether because a prospect failed to meet one
criteria, even while meeting all the others.

In our example we give more weight to single men, but a married
man with a high income and an aging car who is close to forty
would be well worth contacting. And we really don't want to

-2-










eliminate those whose fortieth birthday has already passed. With
the ordinary capabilities of microcomputer DBMS's we would
probably end up being less strict, less creative in setting up our
'view' of the data - and therefor less efficient.

With Intelli-Trieve, on the other hand, we won't 'lose' good
candidates, yet we can still apply all the detailed, complex
criteria we can think up to 'qualify' our leads.

Intelli-Trieve's menu-driven retrieval criteria specification
process allows us to quickly assign each important field's TARGET
VALUE and (for numeric and date fields) a RANGE within which we
are seeking a MATCH. We also specify a MATCH VALUE to be added to
a record's 'weight' if that record's field falls within our
desired range; and a NO-MATCH VALUE to be subtracted when the
match fails.

Some of the fields we are including in our selection criteria will
often contain no data. Intelli-Trieve gives you the option of
'disregarding' these 'NULL FIELDS', so as not to unfairly penalize
a record for not meeting the match criteria, when the value is
simply unknown. Intelli-Trieve accomplishes this in a manner that
also avoids favoring records with many 'nulls'. The result is a
'weighted qualification percentage' based on potential and actual
scores (adjusted for nulls as specified).

It is up to you to decide how much weight to give to each field's
meeting, or not meeting, your selection criteria. In our sports
car sales example, we might weight each of our criteria as
follows:

FIELD FIELD TARGET MATCH NO-MATCH DISREGARD
NAME TYPE VALUE VALUE VALUE NULLS?
----- ----- ------ ----- -------- ---------

SEX CHAR 'M' 500 500 NO

MARRIED LOG .F. 100 50 N/A

INCOME NUM > 40,000 100 200 NO

INCOME NUM > 60,000 200 0 YES

BORN DATE 7/1/47 - 9/30/47 400 50 NO

LAST_CAR DATE < 7/1/85 200 50 YES

CAR_COST NUM > 8,000 100 0 YES

FINANCED LOG .F. 200 0 N/A

Notice that Intelli-Trieve allows more than one criteria per
field, as in the case of INCOME which awards 100 'points' for
incomes greater than $40,000, and an additional 200 if income is

-3-










greater than $60,000. Income below $40,000 results in a point
'penalty' of 200.

Consider the sampling of database records below:

REC# SEX MARRIED INCOME BORN LAST_CAR CAR_COST FINANCED
---- --- ------- ------ -------- -------- -------- --------

1 M .T. 25,000 01/23/52 04/01/86 9,500 .T.

2 F .F. 37,000 / / / / 0

3 M .F. 66,500 11/28/44 05/01/83 11,400 .T.

4 M .T. 68,000 08/23/47 04/15/82 23,500 .F.

5 M .F. 41,000 11/25/46 / / 0

Although none of the records meets all the desired criteria,
several are very 'acceptable' as defined by our selection
criteria. Record #4, for instance, is the best match, even though
he is married. Record #3 is a good fit as well, although he
misses on age, and he financed his last car. And while our data
for record #5 is incomplete, what data we do have shows that he
might be a good prospect. The other two records are less hopeful.

Intelli-Trieve will create a new copy of the database, in which
record #4 will appear first, followed by records #3, #5, #1, and
#2. This new, weighted order will allow our hypothetical marketing
effort to be most fruitful and efficient.

How you use Intelli-Trieve depends on your information needs and
the data you have to work with, limited only by your imagination
and creativity.

INSTALLING AND USING INTELLI-TRIEVE

Intelli-Trieve is simplicity itself to install. Simply copy
IT.EXE to a diskette or to your hard drive, type 'IT' at the DOS
prompt, and the INTELLI-TRIEVE MENU appears.

It is a standard 'light-bar' menu; use the up/down arrows to
highlight the function you want to use, and then press RETURN.

The first menu selection 'REGISTRATION' will give you information
on how to become a registered user of Intelli-Trieve, and how
registration will benefit you. 'REGISTRATION' will also tell you
more about the user survey.

The second menu selection, 'SPECIFY RETRIEVAL CRITERIA' allows you
to identify the database you wish to use, and the name of the
specification criteria file you wish to create. For more
information, select this option and press F1 for more help.


-4-










The third menu selection, 'PERFORM WEIGHTED RETRIEVAL', creates a
new database with the fields in weighted order, based on the
selection criteria you previously saved in a criteria file. For
more information, select this option and press F1 for more help.

The fourth menu selection, 'HELP', displays this tutorial. The
final option, 'QUIT', returns to the DOS prompt or the program
that invoked Intelli-Trieve.

SPECIFYING RETRIEVAL CRITERIA

In order to create retrieval criteria you must tell Intell-Trieve
which database is to be used. You must also name the criteria
file to be created. The criteria file must not already exist.

After naming the database and criteria files, a screen appears
which allows you to name, in turn, each field which you want to
include in the weighted retrieval calculations. Numeric, Date,
Character, and Logical fields are allowed. You may specify
multiple, separate criteria for any individual field. Up to 128
field specifications are allowed.

Entering a valid field name displays a screen for that type of
field. The information required for a field specification depends
somewhat on the type of field, but many of the requirements are
the same for all field types.

INFORMATION REQUIRED FOR ALL FIELD TYPES

MATCH VALUE is the number of 'points' awarded an individual record
for 'matching' the criteria set. It can be any positive number or
zero. The particular value assigned is up to you, and works in
conjunction with all the other MATCH and NO-MATCH VALUES assigned
to other fields.

NO-MATCH VALUE is the number of 'points' an individual record is
penalized for not meeting the criteria specified. It can be any
positive number or zero.

If you answer 'Y' to the question, 'DISREGARD NULL FIELDS?', an
adjustment will be made to the 'weighted qualification percentage'
of any records for which the field contains a zero (numeric), or
blanks (date and character). Logical fields are never 'null', and
therefor cannot be 'disregarded'. This technique avoids unfairly
penalizing records for which information was simply not available.
More detail on 'nulls' is contained in the Intelli-Trieve tutorial
available from the main menu.

NUMERIC FIELDS

When specifying retrieval criteria for a numeric field, enter the
TARGET VALUE (what you're looking for), and the LOWER and UPPER
LIMITS OF THE RANGE. If all three are equal, then an exact match


-5-










is required. Otherwise, a value between the lower and upper
limits, inclusive, will constitute a match.

DATE FIELDS

When specifying retrieval criteria for a date field, enter the
TARGET VALUE (the date you're looking for), and the LOWER and
UPPER LIMITS OF THE RANGE. If all three are equal, then an exact
match is required. Otherwise, a date falling between the lower and
upper limits, inclusive, will constitute a match.

LOGICAL FIELDS

For logical fields you need only specify which constitutes a
match, 'True (T)' or 'False (F)', and what the MATCH and NO-MATCH
VALUES are.

CHARACTER FIELDS

For character fields you may specify any word or phrase up to 20
characters long. Intelli-Trieve matches the word or phrase with
the contents of the character field without regard to whether the
letters are in upper or lower case; the matching is NOT case-
sensitive.

FUNCTION KEYS

While specifying criteria for an individual field, the ESC key
will cancel the individual field specification. When pressed at
the field selection screen, the ESC key cancels the entire
retrieval specification, and does not save any work.

F3 saves the retrieval criteria to the assigned file and returns
to the main menu.

F1 displays help screens.

PERFORMING WEIGHTED RETRIEVAL

When you have completed specifying the retrieval criteria for your
database you are ready to tell Intelli-Trieve to perform the
weighted retrieval calculations and produce a new copy of your
database in weighted order.

You are asked to name the CRITERIA FILE which contains the
retrieval specifications you have created. Include the drive and
path if different from the the current DOS defaults. Intelli-
Trieve will verify that the file exists, and that it is valid.

You are also asked for the name, and optional drive and path, of
the output database to be created. The output database must use
the extension '.DBF'. If the file already exists it will be
overwritten. DO NOT use your database file as Intelli-Trieve's
output database - if you wish to have an output database of the

-6-










same name as your input database, specify a different path or
drive, in order to guarantee that Intelli-Trieve's weighted
retrieval output will be a sorted COPY of your database.

Intelli-Trieve will display messages on the screen as each step in
the weighted retrieval process is performed. When the process is
complete a message at the top of the screen will confirm that a
new database has been created. Use your DBMS to access the
information in it.

The amount of time required to perform weighted retrieval depends
on a number of factors, including the speed of your computer and
disk drive, how large your database is, and how complex your
retrieval criteria are. The more complex the retrieval criteria
the longer the retrieval will take.

THIS IS A SHAREWARE PROGRAM

The Shareware concept, under which this program is distributed,
depends on the direct support of those who use the software. You
are asked to contribute $19 to become a registered user of
Intelli-Trieve.

Your support of this product helps ensure that it will keep up
with changes to dBASE and that enhancements will be made to
increase its usefulness.

As part of Intelli-Trieve's further development process,
registered users are asked to complete the short questionnaire
which follows and enclose it with your registration check. Your
responses will help guarantee that future releases of this program
meet the needs of its registered users.

SEND TO THE ADDRESS BELOW:

-Your $19 contribution (check payable to 'Richard Ray Viets')
-Your name, address and phone number
-Your completed questionnaire

RICHARD RAY VIETS
INTELLI-TRIEVE
16 OLDFIELD FARMS
ENFIELD, CT 06082












-7-










INTELLI-TRIEVE USER QUESTIONAIRE

Rank the following possible enhancements to Intelli-Trieve in
their order of usefulness to you (1 for most useful, 7 for least):

____ Modify an existing criteria file

____ Onscreen browse of records in weighted order without
creating a sorted copy of the database

____ Create a copy of the database in weighted order, but only
copy a specified number of the 'best' records

____ Display the weighted qualification percentage (the record's
'score') onscreen

____ Generate a printable text summary of the criteria specified

____ Specify criteria ranges in percentages or +/- units

____ Specify date criteria as day-of-week or month

COMMENTS:


 December 17, 2017  Add comments

Leave a Reply