CALCULATING RELIABILITY SCORES FROM WORD SPLITTING

• US 20110208723A1
• Filed: 09/30/2010
• Published: 08/25/2011
• Est. Priority Date: 02/19/2010
• Status: Active Grant

• Current Assignee: Go Daddy Operating Company LLC (GoDaddy, Inc.)
• Original Assignee: Go Daddy Group Incorporated (GoDaddy, Inc.)
• Inventors: Glick, Gregory, Lutwitze, Patrick, Schmucker, Doug, Nicks, Paul, Belina, Jeff
• Granted Patent: US 8,706,728 B2
• Time in Patent Office: Days
• US Class Current: 707/723
• CPC Class Codes: G06F 16/3338 Query expansion
• CPC Class Codes: G06Q 10/10 Office automation; Time management
• CPC Class Codes: G06Q 30/0206 Price or cost determination based on market factors
• CPC Class Codes: G06Q 30/0283 Price estimation or determination
• CPC Class Codes: H04L 2101/30 Types of network names

Abstract

Systems and methods of the present invention provide for the word splitting and reliability score for an entered character string. A list of keywords may be extracted from the character string entered into a user interface on a client. These keywords may be compared to potential matches in a dictionary database and a reliability score for word splits and keywords strings may be compiled and displayed to the user. The client may also display the reliability score using a plurality of logical groupings within a reliability score process.

Claims

1. A method, comprising the steps of:

a) receiving a word splitting input as a character string via a user interface on a client computer communicatively coupled to a network to be executed by a processor by one or more word splitting software modules on one or more server computers communicatively coupled to the network;

b) calculating a reliability score for each of one or more word splits derived from the word splitting input, via one or more reliability score software modules executed on the server computer, wherein calculating the reliability score comprises the steps of;

i) comparing a plurality of keywords derived from each of the one or more word splits with a plurality of dictionary words in a dictionary database to determine a percentage of the plurality of keywords found in the dictionary database;

ii) for each of the one or more word splits, determining a percentage of characters in the plurality of keywords found in the dictionary database;

iii) scoring a uniqueness of each of the one or more word splits; and

iv) ranking each of the one or more word splits based on how many of the plurality of keywords were derived from each of the one or more word splits; and

c) displaying each of the one or more word splits and the reliability score for each on the user interface on the client.

2. The method of claim 1 further comprising the step of weighting the percentage of the plurality of keywords found in the dictionary database according to whether or not the percentage of the plurality of keywords found in the dictionary database is 100.

3. The method of claim 1 wherein the reliability score for each of the one or more word splits is calculated by multiplying the percentage of the plurality of keywords found in the dictionary database with a first multiplicand, the ranking of the one or more word splits is multiplied by a second mutliplicand, adding these products to the percentage of characters, the uniqueness of the one or more word splits and the rank of the one or more word splits, and dividing this sum by a divisor.

4. The method of claim 1 wherein, if a first plurality score for a first keyword string is the same as a second plurality score for a second keyword string, an n-gram score using an algorithm and a dataset determines relative usage of keywords together with common language.

5. The method of claim 4 wherein n=the number of words.

6. The method of claim 4 wherein the n-gram algorithm is used as a secondary sort after the reliability scores are determined to determine the most reliability of a final keyword split.

7. The method of claim 1 wherein the uniqueness each of the one or more word splits is higher if the one or more word splits has fewer splits.

8. The method of claim 1 wherein the uniqueness of each of the one or more word splits is determined by how many of the one or more word splits contain the same number of keywords.

9. The method of claim 1 wherein the uniqueness of each of the one or more word splits is scored by calculating a percentage derived from the inverse of the number of the one or more word splits that produces that number of keywords.

10. The method of claim 9 wherein the percentage is converted into a whole number to determine a final score.

11. The method of claim 1 wherein a rank for each of the one or more word splits is higher according to fewer splits within each of the one or more word splits.

12. The method of claim 1 wherein ranking each of the one or more word splits further comprises generating a number for each of the one or more word splits representing a count of the number of keywords for each of the one or more word splits.

13. The method of claim 1 wherein ranking each of the one or more word splits further comprises using the number of keywords for each of the one or more word splits as a rank, wherein the lowest number of keywords is determined to be the highest rank.

14. The method of claim 13 wherein a rank is determined by calculating a percentage derived from the inverse of the rank assigned to the number of keywords.

15. The method of claim 14 wherein the percentage is converted into a whole number to determine a final ranking score.

16. The method of claim 1 wherein the reliability score is used in the context of a spellcheck program.

17. The method of claim 1 wherein the reliability score is used in the context of a search engine program.

18. The method of claim 1 wherein the reliability score is used in the context of a domain setup program.