EP0433964A2
Expired / Lapsed
EP
⏱️ 0.0 years remaining
Name resolution in a directory database
Assignee: Bull HN Information Systems Inc
•
Filing: 1990-12-18
•
Granted: —
•
Expires: 2010-12-18
Abstract
Abstract The present invention provides a method for resolving a common name purported to identify a database entry with the database entry names stored in the index. The purported and entry names each comprise a first field, a second field, and a third field, with at least the first field including at least one word. Purported names are provided to the name resolution process in a normalized form referred to as the "Common Name" format. Entry names are stored in a similar normalized form. The name resolution process begins by determining an exact match score and initializing count fields. The process then constructs a set of all index entries having first (surname) fields matching the first (surname) field of the purported word. The process then selects an initial "guess" at the surname component of the purported name and identifies all index entries having either (a) surnames equal to the last word of the purported name "surname" field or (b) multiple word surnames whose last word equals the last word of the purported name "surname" field. Then, for each of the entry names in the set, the process compares the first fields of the entry name and purported name and, if the first fields of the purported and entry names do not match, selects another entry name from the set. The process then selects a second field from the purported name and compares the second fields of the purported and entry names. In this comparison, the order of words in the second fields is immaterial, that is, only the number of matching words in the fields is considered. The process selects a third (personal name) field from the purported name and performs the following sequence of steps for each word in the third field of the purported name. The process compares the selected word of the purported name with the word in the same position in the third field of the entry name. If the words match, the process selects the next word in the third (personal name) field of the purported name and returns to the start of this sequence. If the words do not match, but a word exists in the third field of the entry name that is comparable to the word selected from the third field of the purported name, the process compares the initial characters of the words in the third fields of the purported and entry names. An alternate or additional embodiment of the process of the present invention also provides for an additional comparison for alternate or alias names in the third (personal name) field. The final step of the process then determines a comparison score for the degree of similarity between the purported and entry names and selects the best match.
Inventors
- 👤 Ronald Everett Dann
Bibliographic Details
Application Number
—
Jurisdiction
EP
Data Source
google_patents
Last Synchronized
2026-11-09 14:44
CPC Classifications
H; H04; H04Q; H04Q3/00; H04Q3/42; H04Q3/47; G; G06; G06F; G06F16/00; G06F16/90; G06F16/903; G06F16/90335; G06F16/90348; H04M; H04M3/00; H04M3/42; H04M3/487; H04M3/493; H04M3/4931; Y; Y10; Y10S; Y10S707/00; Y10S707/99931; Y10S707/99933; Y10S707/99935
Claims Analysis
Total Claims: 19
Claim 1
Independent Claim
476 words
In an information processing system having a database and an index of entries in the database, a method for resolving a common name purported to identify a database entry with the database entry names stored in the index, wherein each purported name and each entry name comprises a surname field, a generation and awards field, and a personal name field and wherein at least the surname field includes at least one word, comprising the steps of: (a) determining a match score representing an exact match between the purported name and an entry name, (b) initializing a set of count values wherein each count value represents the similarity between a field of a purported name and a corresponding field of an entry name, (c) selecting an surname field from the purported name, (d) constructing a set of all entry names having the same surname field as the purported surname field, and (e) for each of the entry names in the set, (f) comparing the surname field of the entry name with the surname field of the purported name, (g) if the surname fields of the purported and entry names do not match, returning to step (e) and selecting another entry name from the set, (h) if the surname fields of the purported and entry names match, incrementing the corresponding count values, (i) selecting a generation and awards field from the purported name, (j) comparing the generation and awards fields of the purported and entry names, and, if matches are found in the generation and awards fields, incrementing the corresponding count values, (k) selecting a personal name field from the purported name and, for each word in the personal name field of the purported name, (l) comparing the selected word of the purported name with the word in the same position in the personal name field of the entry name, (m) if no comparable word exists in the personal name field of the entry name, going to step (r), (n) if the words match, incrementing the corresponding count values and go to step (r), (o) if the words do not match, comparing the initial characters of each of the words in the personal name fields of the purported and entry names, and, (p) if the words of the personal name field of one name are an abbreviated form of the words of the other, incrementing the corresponding count values and going to step (r), (q) if the words of the personal name fields of neither of the names is an abbreviated form of the other, selecting another entry name from the set and returning to step (e), (r) determining a comparison score for the degree of similarity between the purported and entry names as a weighted sum of the count values, and (s) identifying an entry name of the set as most likely corresponding to the purported name.
Claim 2
Dependent Claim
464 words
In an information processing system having a database and an index of entries in the database, a method for resolving a common name purported to identify a database entry with the database entry names stored in the index, wherein each purported name and each entry name comprises a first field, a second field, and a third field and wherein at least the first field includes at least one word, comprising the steps of: (a) determining a match score representing an exact match between the purported name and an entry name, (b) initializing a set of count values wherein each count value represents the similarity between a field of a purported name and a corresponding field of an entry name, (c) selecting an first field from the purported name, (d) constructing a set of all entry names having the same first field as the purported first field, (e) for each of the entry names in the set, (f) comparing the first field of the entry name with the first field of the purported name, (g) if the first fields of the purported and entry names do not match, returning to step (e) and selecting another entry name from the set, (h) if the first fields of the purported and entry names match, incrementing the corresponding count values, (i) selecting a second field from the purported name, (j) comparing the second fields of the purported and entry names, and, if matches are found in the second fields, incrementing the corresponding count values, (k) selecting a third field from the purported name and, for each word in the third field of the purported name, (l) comparing the selected word of the purported name with the word in the same position in the third field of the entry name, (m) if no comparable word exists in the third field of the entry name, going to step (r), (n) if the words match, incrementing the corresponding count values and go to step (r), (o) if the words do not match, comparing the initial characters of each of the words in the third fields of the purported and entry names, and (p) if the words of the third field of one name are an abbreviated form of the words of the third field of the other name, incrementing the corresponding count values and going to step (r), (q) if the words of the third fields of neither of the names is an abbreviated form of the other, selecting another entry name from the set and returning to step (e), (r) determining a comparison score for the degree of similarity between the purported and entry names as a weighted sum of the count values, and (s) identifying an entry name of the set as most likely corresponding to the purported name.
Claim 3
Dependent Claim
96 words
The method for name resolution of claim 2, wherein: the database entries contain information relating to persons and the purported and entry names are names referring to persons, the first fields of the purported and entry names are surname fields containing information relating to the surnames of the persons, the second fields of the purported and entry names are generation and awards fields containing generation and awards information relating to the persons, and the third fields of the purported and entry names are personal name fields containing information relating to the personal names of the persons.
Claim 4
Dependent Claim
17 words
The method for name resolution of claim 3 wherein a surname may be comprised of multiple words.
Claim 5
Dependent Claim
21 words
The method for name resolution of claim 4 wherein two or more words of a multiple word personal may be hyphenated.
Claim 6
Dependent Claim
23 words
The method for name resolution of claim 3, wherein the personal name fields contain information relating to the person's first and middle names.
Claim 7
Dependent Claim
18 words
The method for name resolution of claim 6 wherein the personal name fields may contain alternate first names.
Claim 8
Dependent Claim
24 words
The method for name resolution of claim 6 wherein the personal name fields may contain the initials of the person's first and middle names.
Claim 9
Dependent Claim
18 words
The method for name resolution of claim 6 wherein a first name may be comprised of multiple words.
Claim 10
Dependent Claim
194 words
The method for name resolution of claim 2, wherein: the first field selected in step (c) is an initial guess at a purported name first field, and step (e) comprises the further steps of (e1) examine the first field of an entry name, (e2) if the first field of the entry name contains a single word, increment a corresponding count value, (e3) if the first field of the entry name contains n words, where n is greater than 1, (e4) form a second guess at a purported name first field by prefixing the (n-1) words following the first word of the word of the initial guess at a purported name first field to the first word, (e5) compare the second guess at the purported name first field with the first field of the entry name, (e6) if the second guess at the purported name first field and the first field of the entry name do not match, select another entry name from the set and return to step (e1), (e7) if the second guess at the purported name first field and the first field of the entry name match, increment a corresponding count value.
Claim 11
Dependent Claim
62 words
The method for name resolution of claim 2, wherein the comparison of the second fields of the purported and entry names is performed by the steps of: (j1) for each word in the second field of the purported name, examine the second field of the entry name for an identical word at any position in the second field of the entry name.
Claim 12
Dependent Claim
40 words
The method for name resolution of claim 11, wherein step (j) further comprises the step of: (j2) if the second fields of the purported and entry names contain exactly the same words, in any order, increment a corresponding count value.
Claim 13
Dependent Claim
200 words
The method for name resolution of claim 2, wherein the comparison of the initial characters of the third fields of the purported and entry names of step (o) comprises the further steps of: (o1) expressing the third field of the entry name according to the expression ir [-ir] m , where i is the initial character of a substring of the third field of the entry name that contains no hyphen characters, r is the possibly null remainder of that substring, and m is an integer =>0, and (o2) determining whether the third field of the purported name has the form i [i | -i] n where n<=m, wherein the third field of the purported name is an initial form of the third field of the entry name if the third field of the purported name is of this form, and (o2) if the third field of the purported name is not an initial form of the third field of the entry, repeating steps (o1) and (o2) with the third fields of the entry and purported names reversed to determine whether the third field of the entry name is an initial form of the third field of the purported name.
Claim 14
Dependent Claim
89 words
The method for name resolution of claim 2 wherein the third fields of the entry names may contain alternate names and wherein step (p) further comprises the steps of: (p1) if the words of the third fields of neither of the names is an abbreviated form of the other, selecting an alternate name word corresponding to the present third field word of the entry name, and (p2) repeating steps (m) through (p) for each successive alternate name word corresponding to the present third field word of the entry name.
Claim 15
Dependent Claim
106 words
The method of name resolution of claim 2, wherein step (s) is comprised of the steps: (s1) if the comparison score is equal to the exact match score, terminate the name resolution process and provide the entry name as corresponding to the purported name, (s2) if the comparison score is not equal to the exact match score, record the comparison score, select another entry name from the set, and return to step (e), and (s3) if there are no further entry names to be resolved in the set, terminate the name resolution and provide at least the entry name most likely corresponding to the purported name.
Claim 16
Dependent Claim
34 words
The method of name resolution of claim 3, wherein each purported name is of the form, and in the order listed, a [personal name field], a [surname field], and a [,generation and awards field].
Claim 17
Dependent Claim
49 words
The method of name resolution of claim 16, wherein step (c) of selecting a first field from the purported name further includes the step of: selecting from the purported name the last word before any comma or, if there is no comma, the last word in the purported name.
Claim 18
Dependent Claim
34 words
The method of name resolution of claim 16, wherein step (i) of selecting a second field of the purported name includes the step of: selecting the words following the comma in the purported name.
Claim 19
Dependent Claim
39 words
The method of name resolution of claim 16, wherein step (k) of selecting a third field of the purported name includes the step of: constructing the third field from all words in the purported name preceding the first field.
References Cited (Backward Citations)
No citation records indexed.
USPTO Annuity & Maintenance Windows
Maintenance schedule is calculated for granted US utility patents.