Filtering phrases for an identifier

• US 8,423,349 B1
• Filed: 01/13/2009
• Issued: 04/16/2013
• Est. Priority Date: 01/13/2009
• Status: Active Grant

• Current Assignee: Amazon Technologies, Inc. (Amazon.com, Inc.)
• Original Assignee: Amazon Technologies, Inc. (Amazon.com, Inc.)
• Inventors: Huynh, Steve, Oates, Isaac, Jesensky, James, Vaidya, Vinay
• Primary Examiner(s): Smits, Talivaldis Ivars
• Assistant Examiner(s): ROBERTS, SHAUN A
• Application Number:
  US12/353,120
• Time in Patent Office: 1,554 Days
• Field of Search: 705/74, 705/44, 705/67, 704/1, 704/9, 726/6
• US Class Current: 704/9
• CPC Class Codes: G06F 40/56 Natural language generation

Abstract

Techniques for generating and providing phrases are described herein. These techniques may include analyzing one or more sources to generate a first corpus of phrases, each of the phrases for use as an identifier and/or for association with a user. The first corpus of phrases may be generated based on characteristics associated with phrases. Once a first corpus of phrases has been generated, these phrases may be filtered to define a second corpus of phrases. Phrases of this second corpus may then be suggested to one or more users. In some instances, the phrases suggested to a particular user are personalized to the user based on information previously known about the user or based on information provided by the user.

Claims

1. One or more computer-readable media storing computer-executable instructions that, when executed on one or more processors, perform acts comprising:

analyzing multiple corpuses of text to generate a first corpus of phrases, each of the phrases comprising at least two grammatically-correct words;

determining, for each phrase of the first corpus of phrases;

(i) a commonality of the phrase;

(ii) a commonality of each word of the phrase;

(iii) a number of syllables in the phrase, and (iv) a number of words in the phrase, the commonality of the phrase indicating a frequency of an occurrence of the phrase;

scoring each phrase of the first corpus of phrases based on the commonality of the phrase, the commonality of each word of the phrase, the number of syllables in the phrase, and the number of words in the phrase;

filtering the first corpus of phrases to define a second corpus of phrases comprising fewer phrases than the first corpus of phrases, wherein the filtering comprises;

removing phrases of the first corpus of phrases based at least in part on scoring of the phrases;

removing phrases of the first corpus of phrases that include words appearing on a predetermined blacklist; and

removing phrases of the first corpus of phrases that comprise specified part-of-speech combinations;

providing phrases of the second corpus of phrases to a device for selection; and

responsive to receiving a selection of a phrase, associating the selected phrase with the user.

2. One or more computer-readable media storing computer-executable instructions that, when executed on one or more processors, perform acts comprising:

generating a first corpus of phrases each comprising a string of alphanumeric characters;

filtering the first corpus of phrases to define a second corpus of phrases comprising fewer phrases than the first corpus of phrases; and

associating a phrase of the second corpus of phrases with at least one of a shipping address associated with a user, a device associated with the user or location data in a server associated with the user.

3. One or more computer-readable media as recited in claim 2, wherein phrases of the first corpus of phrases each comprise one or more words, and further storing computer-executable instructions that, when executed on the one or more processors, perform acts comprising:

determining, for each phrase of the phrases comprising one or more words, a part-of-speech for each of the words of the phrase; and

determining, based on the determined parts-of-speech of the words, a part-of-speech combination for each of the phrases comprising one or more words;

and wherein the filtering of the first corpus of phrases comprises filtering out phrases based at least in part on the determined part-of-speech combinations.

4. One or more computer-readable media as recited in claim 2, wherein phrases of the first corpus of phrases each comprise one or more words, and further storing computer-executable instructions that, when executed on the one or more processors, perform acts comprising:

determining, for each phrase of the phrases comprising one or more words;

(i) a commonality of the phrase;

(ii) a commonality of each word of the phrase;

(iii) a readability of the phrase;

(iv) a number of words in the phrase, or (v) a source of the phrase; and

scoring each phrase of the phrases comprising one or more words based on the commonality of the phrase, the commonality of each word of the phrase, the readability of the phrase, the number of words in the phrase, or the source of the phrase;

and wherein the filtering of the first corpus of phrases comprises filtering out a phrases based at least in part on the scoring of the phrases.

5. One or more computer-readable media storing computer-executable instructions that, when executed on one or more processors, perform acts comprising:

generating a first corpus of phrases each comprising at least two grammatically-correct words;

filtering the first corpus of phrases to define a second corpus of phrases, wherein the filtering the first corpus of phrases comprises;

determining a commonality of words of the multiple phrases of the first corpus of phrases; and

filtering out phrases of the first corpus of phrases to define a second corpus of phrases based at least in part on the determined commonality of the words of the multiple phrases;

outputting phrases of the second corpus of phrases to each of multiple users;

receiving selections of phrases from each of the multiple users; and

responsive to the receiving of the selection, associating a selected phrase with each respective user of the multiple users.

6. One or more computer-readable media as recited in claim 5, wherein the filtering of the first corpus of phrases is further based at least in part on part-of-speech combinations of the phrases of the first corpus of phrases.

7. One or more computer-readable media as recited in claim 5, wherein the filtering of the first corpus of phrases further comprises:

comparing phrases or words of phrases of the first corpus of phrases to a blacklist of words, a blacklist of phrases, or a blacklist of words and phrases; and

removing;

(i) phrases of the first corpus of phrases that appear on the blacklist of words, a blacklist of phrases, or a blacklist of words and phrases, or (ii) phrases of the first corpus of phrases having words that appear on the blacklist of words and/or phrases.

8. One or more computer-readable media as recited in claim 5, wherein the filtering of the first corpus of phrases further comprises:

comparing words of phrases of the first corpus of phrases to a list of words; and

removing phrases of the first corpus of phrases that have words that do not appear on the list of words.

9. One or more computer-readable media as recited in claim 5, wherein the filtering of the first corpus of phrases further comprises:

determining phrases that have previously been associated with users; and

filtering out phrases of the first corpus of phrases that have previously been associated with the users.

10. One or more computer-readable media as recited in claim 5, wherein the filtering of the first corpus of phrases further comprises:

determining phrases that have previously been associated with users;

determining homophones of the phrases that have previously been associated with users; and

filtering out;

(i) phrases of the first corpus of phrases that have previously been associated with the users, and (ii) the determined homophones of the phrases that have previously been associated with the users.

11. One or more computer-readable media as recited in claim 5, wherein the filtering of the first corpus of phrases further comprises:

determining a readability of each of multiple phrases of the first corpus of phrases; and

filtering out a phrases of the first corpus of phrases based at least in part on the determined readability of the multiple phrases.

12. One or more computer-readable media as recited in claim 11, wherein the readability of the multiple phrases is determined based on a Flesch-Kincaid Index or a Coleman-Liau Index.

13. One or more computer-readable media as recited in claim 5, wherein the filtering of the first corpus of phrases further comprises:

determining a word count of each of multiple phrases of the first corpus of phrases; and

filtering out phrases of the first corpus of phrases based at least in part on the determined word count of the multiple phrases.

14. One or more computer-readable media as recited in claim 5, wherein the filtering of the first corpus of phrases further comprises:

determining a source that generated each of multiple phrases of the first corpus of phrases; and

filtering out phrases of the first corpus of phrases based at least in part on the determined sources that generated each of the multiple phrases.

15. A system comprising:

one or more processors;

memory;

a phrase-generation module, stored in the memory and executable on the one or more processors to generate multiple phrases for association with one or more aspects of user accounts associated with corresponding users, the one or more aspects of the user accounts for executing a payment transaction;

a phrase-filtering module, stored in the memory and executable on the one or more processors to filter out one or more of the generated phrases based at least in part on a commonality of one or more words of a phrase; and

a phrase-association module, stored in the memory and executable on the one or more processors to associate one phrase that has not been filtered out with an aspect of a user account for executing a payment transaction.

16. A system as recited in claim 15, wherein the phrase-filtering module comprises a part-of-speech filter that is executable on the one or more processors to filter out generated phrases based on part-of-speech combinations of the generated phrases.

17. A system as recited in claim 15, wherein the phrase-filtering module filters phrases based on a perceived interestingness of each of the multiple phrases.

18. A system as recited in claim 15, wherein the phrase-filtering module filters out:

(i) phrases that have previously been associated with users, and (ii) phrases that sound like phrases that have previously been associated with users.

19. A system as recited in claim 15, wherein the phrase-filtering module filters out phrases based at least in part on the generation of multiple phrases by the phrase-generation module.

20. A method implemented at least in part by a computing device, the method comprising:

generating a first corpus of phrases, each of the phrases comprising at least two grammatically-correct words;

filtering the first corpus of phrases to define a second corpus of phrases; and

outputting phrases of the second corpus of phrases for use as identifiers of a user;

receiving, from a device, a selection of an outputted phrase; and

responsive to the receiving of the selection of the outputted phrase, associating the selected phrase with an account of the user, at least a portion of the selected phrase being generated by the computing device.

21. A method implemented at least in part by a computing device, the method comprising:

receiving a first corpus of phrases, wherein individual phrases comprise at least one word having an associated part-of-speech;

identifying, for one or more phrases of the first corpus of phrases, a part-of-speech combination based on the associated parts-of-speech of the words of the phrase; and

determining, by the computing device, account information associated with a user;

based at least in part on identified part-of-speech combinations and the determined account information, outputting a portion of the phrases as identifiers of the user.

22. A method as recited in claim 21, further comprising mining multiple corpuses of text to generate the first corpus of phrases.

23. A method as recited in claim 21, wherein the outputted portion of the phrases comprises a first portion, and further comprising filtering out a second portion of the first corpus of phrases based on associated parts-of-speech of the phrases before outputting the first portion of phrases.

24. A method as recited in claim 21, wherein at least one phrase comprises at least one word having multiple associated parts-of-speech, and wherein the identifying for the at least one phrase comprises identifying multiple part-of-speech combinations based on the associated parts-of-speech of the words of the at least one phrase.

25. A method implemented at least in part by a computing device, comprising:

comparing a part-of-speech list to a part-of-speech combination for a set of multiple phrases;

based at least in part on the comparing, identifying one or more phrases of the multiple phrases for a suggestion; and

suggesting the identified one or more phrases to a device;

receiving, from the device, a selection of an outputted phrase; and

responsive to the receiving of the selection of the outputted phrase, associating the selected phrase with a payment instrument of the user, at least a portion of the selected phrase being generated by the computing device.

26. A method as recited in claim 25, wherein each of the multiple phrases comprise at least two grammatically-correct words.

27. A method as recited in claim 25, wherein the part-of-speech list specifies part-of-speech combinations for exclusion, and further comprising filtering out phrases having part-of-speech combinations specified on the part-of-speech list.

28. A method as recited in claim 25, wherein the part-of-speech list specifies part-of-speech combinations for inclusion, and further comprising filtering out phrases having part-of-speech combinations that are not specified on the part-of-speech list.

29. One or more computer-readable media storing computer-executable instructions that, when executed on one or more processors, perform acts comprising:

receiving a first corpus of phrases, each of the phrases comprising at least one word;

determining, for each phrase of the first corpus of phrases;

(i) a commonality of the phrase;

(ii) a commonality of each word of the phrase;

(iii) a readability of the phrase;

(iv) a number of words in the phrase, or (v) a source of the phrase;

scoring each phrase of the first corpus of phrases based on the commonality of the phrase, the commonality of each word of the phrase, the readability of the phrase, the number of words in the phrase, or the source of the phrase; and

based at least in part on the scoring of the phrases, filtering out at least a portion of the first corpus of phrases to define a second corpus of phrases;

outputting phrases of the second corpus of phrases to a user; and

responsive to receiving a selection of a phrase by the user, associating the selected phrase with at least one of a shipping address, a device or location data in a server that is associated with the user.

30. One or more computer-readable as recited in claim 29, further storing computer-executable instructions that, when executed on the one or more processors, perform an act comprising analyzing multiple corpuses of text to generate the first corpus of phrases.

31. One or more computer-readable as recited in claim 29, wherein the source of the phrase comprises a manner in which the phrase was generated.

32. A method implemented at least in part by a computing device, comprising:

generating a first corpus of phrases, each of the phrases comprising at least one word;

scoring each phrase of the first corpus of phrases based on a perceived interestingness of the phrase and on a commonality of one or more words of a phrase;

based at least in part on the scoring of the phrases, filtering out at least a portion of the first corpus of phrases to define a second corpus of phrases, and

outputting phrases of the second corpus of phrases for use as identifiers of a user.

33. A method as recited in claim 32, further comprising outputting phrases of the second corpus of phrases for association with users and payment instruments of the users.

34. A method implemented at least in part by a computing device, comprising:

generating a first corpus of phrases, each of the phrases comprising a string of alphanumeric characters;

scoring each phrase of the first corpus of phrases based on multiple parameters;

based at least in part on the scoring of the phrases, filtering out at least a portion of the first corpus of phrases to define a second corpus of phrases; and

associating a phrase of the second corpus of phrases with at least one of a shipping address associated with a user, a device associated with a user or location data in a server associated with the user.

35. A method as recited in claim 34, wherein each of phrases comprises at least one grammatically-correct word.

36. A method as recited in claim 34, wherein the multiple parameters comprise one or more of a commonality of the phrase, a commonality of a word of the phrase, a readability of the phrase, a number of words in the phrase, or a source of the phrase.