Method of standardizing address data6438546Abstract A method of standardizing address data in a database using a word dictionary and a pattern dictionary. The method includes the steps of a) breaking up a set of address data into lines; b) breaking up each line into words; c) looking up each word in the word dictionary for identifying the field type of the word; d) forming a line pattern from the field type of the words in the line; e) looking up the line pattern in the pattern dictionary; and f) returning a line pattern to each of the lines in the address data. With each word in each line in a set of address data having a field type assigned thereto, the address components can be easily identified by a machine reading the address data. With the standardized address data, for example, a machine can identify which component of an address is the street name, and which component of a name line is the title of the addressee. Claims What is claimed is: Description TECHNICAL FIELD
N1 Personal Name 1
N2 Personal Name 2
N3 Personal Name 3
F1 Firm/Company
A1 Primary Address
A2 Alternate Address
L1 Last Line/City
S Scan data which is an unidentifiable input line
With the address example given above, the lines are identified as:
N1 Professor and Mrs. Thomas X Johnson
N2 Samuel Connaly, Jr., Esq.
F1 c/o Quimby Imaging Technology, Inc.
A1 123 S. Cross River Pkwy, N.E.
A2 Ste 203-B
L1 East Carrolton, Georgia 31072-0124
B) breaking up each line into words
N1: Professor
and
Mrs
Thomas
X
Johnson
N2: Samuel
Connaly
Jr
Esq
F1: c/o
Quimby
Imaging
Technology
Inc
A1: 123
S
Cross
River
Pkwy
N.E.
A2:
Ste
203-B
L1: East
Carrolton
Georgia
31072
0124
C) Looking up words in a word dictionary The word dictionary, according to the present invention, is used to identify the characteristics of the words, and the characteristics are represented by field types or values. For example, there are 20 type of words, represented by:
Value Type Examples
T State word California, CA, D.C., WI
U Unit designator word Apt., Suite, Ste
R Rural route RR, Hwy
H Highway Contract HC
M Military word CMR, PSC
P PO Box word P.O. Box, Box, POBOX
S Street suffix word Ave, Avenue, Bvld, Pkwy
D Street direction word East, E., N.W.
C Care of word c/o
O Numeric data spelt word Twenty, One
N Numeric street word 5th, 22nd
+ Conjunction word and, &, OR
I Ignored word BLOCK, BLOQUE
(Puerto Rico only)
F Firm word Co, Company, Ltd, LLC
1 Personal name Johnson, Mary
B Financial Words Trustee, Deceased
M Multiple meaning word St (street or saint)
2 Concatenating word La, Des, Las
3 Personal name prefix Ms, Dr, Prof, Professor
4 Personal name suffix Jr, Sr, III, Esq, Ph.D., MD
W word not found in dictionary
In the above example, the values of the words are returned as follows:
Type Word Value
N1:
Title 1 Professor 3
Ignored and +
Title 2 Mrs 3
First Name Thomas 1
Middle Name X W
Surname 1 Johnson 1
Surname 2 --
Suffix 1 --
Suffix 2 --
N2:
Title 1 --
Title 2 --
First Name Samuel 1
Middle Name --
Surname 1 Connaly 1
Surname 2 --
Suffix 1 Jr 4
Suffix 2 Esq 4
It should be noted that, in the Personal Name lines, Surname 2 is the name after the hyphen in a hyphenated surname. For example, in Mrs. Stella Lee-Guggenheim, Guggenheim is categorized as Surname 2.
Type Word Value
F1:
Care of c/o C
Firm name Quimby W
Firm name Imaging W
Firm name Technology W
Firm name Inc F
It should be noted that the breaking up of a Firm Name line is for identification purposes only. The line is broken up so that the words can be checked against a group of firm name words such as Co, Company, LLC, LLP, LTD, Inc., etc., but it is not necessary to use the word types to form a line pattern.
Type Word Value
A1:
Range 123 N
Pre-Dir. S D
Street Name Cross W
Street Name River W
Street Suffix Pkwy S
Post-Dir. N.E. D
A2:
Unit Name Ste U
Unit Number 203-B N
Box Type --
Route Number --
Box Number --
L1:
City Name East W
City Name Carrolton W
State Code Georgia T
Zip Code 31072 N
Plus4 0124 N
D) Identifying the word Pattern in each line L
Line Line pattern
N1: 331W1
N2 1144
F1 --
A1 NDWWSD
A2 UN
L1 WWTNN
It should be noted that the ignored words (type 1) have been dropped from the line patterns. E) standardizing address data In this example, the word pattern in line N1 can be considered non-standard because it contains two titles: Professor and Mrs. It is preferred that the address is rearranged to become:
N1 T1W1 Professor Thomas X Johnson
N2 T1W1 Mrs. Thomas X Johnson
N3 11TT Samuel Connaly, Jr., Esq.
F1 -- Quimby Imaging Technology, Inc.
A1 NDWWS 123 S. Cross River Pkwy, N.E.
A2 UN Suite 203-B
L1 WWTNN East Carrolton, GA 31072-1024
The splitting of the original N1 line to N1 and N2 is called dual-name isolation. Another example of dual-name isolation is to separate the name line of Jane and John Smith into Jane Smith and John Smith. In the last line of the above example, the word "ste" has been replaced by "Suite", and the full state name "Georgia" has been replaced by the abbreviated state name "GA". With the standardized address, it is quite straightforward for a machine to pick out the correct address components based on the values, when required. For example, the city name "East Carrolton" is easily identified by the fact that the city state line has a line pattern of WWTNN, where T is the value of a state name. FIG. 2 illustrates the flowchart of the Blockscan procedure. The Blockscan procedure is used to identify the type of a scanned line in order to indicate whether a line is a Name line, an Address line, etc. The line types are represented by the following values:
A Address line
F Firm line
P Personal name line
C City state line
+ To be concatenated with the line above
- To be concatenated with the line below
U Unknown line
As shown in the flowchart in FIG. 2, the Blockscan procedure includes the following steps: 1) reading a set of address data (50); 2) breaking up the address data into lines (52); 3) going to the bottom data line (54); 4) looking for a state word (56); If a state word is found, go to Step 5; If not, look for a zip word (58); If a zip word is not found, return a "U" and move up a line (60), and repeat Step 4; If a zip word is found, return a "+" and move up a line (62), then look for a state word in the new line (64); If a state word is found in the new line, go to Step 5; If not, return a "U" and move up a line (66), and look for a state word in the new line again (64); 5) looking for words before the state word in the same line (68); If one or more words are found, assume they are city name and go to Step 6; If no word is found, return a "C" and move up a line (72), then look for a street word in the new line (74); If a street word is found, assume state name is located on the street address line and go to Step 8; If a street word is not found, return a "-" and move up a line (76), then go to Step 7; 6) returning a "C" and move up a line (70); 7) looking for a street word in the new line (78); If a street word is found, go to Step 8; If a street word is not found, then return a "U" and move up a line (80); If no more new lines, go to Step 12; Otherwise go back to Step 7; 8) returning an "A" and move up a line (84); 9) looking for a street word again (86); If a street word is found, return an "A" and move up a line (88), then go to Step 10; If a street word is not found, go to Step 10; 10) looking for a firm word (90); If a firm word is found, return an "F" and move up a line (92); If no more new lines, go to Step 12; Otherwise go to Step 11; 11) looking for a name word (96); If a name word is found, return a "P" and move up a line (98); If no more new lines, go to Step 12; Otherwise go back to Step 11; If a name word is not found, return a "U" and move up a line (100); If no more new lines, go to Step 12; Otherwise go back to Step 11; 12) concatenating lines with either "+" or "-" to another line.
Value Scanned line Processes
Example 1:
A 123 Main Street (78, yes)
- Westfield (74, no)
C NJ (64, yes; 68, no)
+ 07090 (56, no; 58, yes)
Example 2:
A 123 Main Street (78, yes)
C Westfield, NJ (64, yes; 68, yes)
+ 07090 (56, no; 58, yes)
Example 3:
A 123 Main Street (78, yes)
- Westfield (74, no)
C NJ 07090 (56,yes; 68, no)
Example 4:
A 123 Main Street (78, yes)
C Westfield, NJ 07090 (56, yes; 68, yes)
The lines in Examples 1-3 will be concatenated, in Step 12, to become the address lines of Example 4. It should be noted that the flowchart shown in FIG. 2 is for illustrative purposes only. It is used to demonstrate the principle of the Blockscan procedure for identifying the line type of each line in an address. The procedure can be varied to include more steps or to skip some steps. For example, according to the flowchart in FIG. 2, the street word is checked only on two lines in Step 7 and Step 5, and the firm word is checked only on one line in Step 10. However, more lines can be checked for street words and firm words if desired. Moreover, when a certain word type is not found in a line, more criteria can be used to identify the line before returning a "U" to it, such as the process (66) of finding a street word. Furthermore, the Blockscan procedure and address standardization procedure can also include the additional step of identifying the country name. The same procedures can be modified for standardizing the address data written in other languages. Although the invention has been described with respect to a preferred process and version thereof, it will be understood by those skilled in the art that the foregoing and various other changes, omissions and deviations in the form and detail thereof may be made without depart from the spirit and scope of this invention.
|
Same subclass Same class Consider this |
||||||||||