Automatic extraction of multilingual dictionary items from non-parallel, multilingual, semi-structured data

• US 9,530,161 B2
• Filed: 02/28/2014
• Issued: 12/27/2016
• Est. Priority Date: 02/28/2014
• Status: Active Grant

• Current Assignee: eBay Inc.
• Original Assignee: eBay Inc.
• Inventors: Sawaf, Hassan
• Primary Examiner(s): ZHU, RICHARD Z
• Application Number:
  US14/194,606
• Publication Number:
  US 20150248400A1
• Time in Patent Office: 1,033 Days
• US Class Current: 1/1
• CPC Class Codes: G06F 16/903 Querying for retrieval from the web G06F16/953
• CPC Class Codes: G06F 40/40 Processing or translation of natural language natural language analysis G06F40/20; semantic analysis G06F40/30
• CPC Class Codes: G06F 40/42 Data-driven translation
• CPC Class Codes: G06F 40/44 Statistical methods, e.g. probability models
• CPC Class Codes: G06F 40/45 Example-based machine translation; Alignment
• CPC Class Codes: G06F 40/49 using very large corpora, e.g. the web
• CPC Class Codes: G06F 40/51 Translation evaluation
• CPC Class Codes: G06F 40/58 Use of machine translation, e.g. for multi-lingual retrieval, for server-side translation for client devices or for real-time translation
• CPC Class Codes: G06N 20/00 Machine learning
• CPC Class Codes: G06Q 30/0601 Electronic shopping [e-shopping]
• CPC Class Codes: G06Q 30/0609 Buyer or seller confidence or verification

Abstract

User interfaces receive a first plurality of user queries and result sets that are in a category, exhibit a constraint, and exhibit user behavior. Also received are a second plurality of user queries and result sets that that are in the category, exhibit the constraint, and exhibit user behavior. The second user queries and results are received either from a plurality of user interfaces coupled to a second system, or from the second system itself. Responsive to detecting that the first plurality of user queries and result sets and the second plurality of user queries and result sets satisfy respective thresholds, a signal indicates that at least one of the first plurality of user queries and at least one of the second plurality of user queries are translations of each other.

Claims

1. A method of machine translation comprising:

receiving at a second publication system, from a first publication system, a first plurality of user queries that are in a first language, that are in a category and that exhibit a constraint, and a first plurality of result sets, respective ones of which correspond to respective ones of the first plurality of user queries and that exhibit a first group of implicit user behavior with respect to the first plurality of user result sets;

receiving, at the first publication system, from the second publication system, a second plurality of user queries that are in a second language, that are in the category and that exhibit the constraint, and a second plurality of result sets, respective ones of which correspond to respective ones of the second plurality of user queries and that exhibit a second group of implicit user behavior with respect to the second plurality of result sets;

detecting, by a user behavior module of a machine translation application in the second publication system, implicit user behavior in the first group of implicit user behavior that indicates that at least some of the first plurality of user queries are satisfied by respective ones of the first plurality of result sets;

detecting, by a user behavior module of a machine translation application in the first publication system, implicit user behavior in the second group of implicit user behavior that indicates that at least some of the second plurality of user queries are satisfied by respective ones of the second plurality of result sets;

detecting that the at least some of the first plurality of user queries and the at least some of the second plurality of user queries satisfy respective thresholds; and

responsive to detecting that the at least some of the first plurality of first user queries and the at least some of the second plurality of user queries satisfy respective thresholds, providing an indication to at least one of the machine translation application in the second publication system or the machine translation application in the first publication system that at least one of the first plurality of user queries and at least one of the second plurality of user queries are likely to be translations of each other.

2. The method of claim 1 wherein the constraint comprises one of aspects, a bar code or a seller.

3. The method of claim 1 wherein the first group of implicit user behavior and the second group of implicit user behavior comprise one of positive behavior, potentially negative behavior or clearly negative behavior.

4. The method of claim 1 wherein the respective thresholds comprise a first percentage of the plurality of the first user queries that is satisfied by result sets responsive to the plurality of first user queries, and a second percentage of the second plurality of user queries that is satisfied by result sets responsive to the plurality of second user queries, and the satisfaction is based on one of positive user behavior, potentially negative user behavior, or clearly negative user behavior.

5. One or more computer-readable hardware storage device having embedded therein a set of instructions which, when executed by one or more processors of a computer, causes the computer to execute operations comprising:

receiving at a second publication system, from a first publication system, a first plurality of user queries that are in a first language, that are in a category and that exhibit a constraint, and a first plurality of result sets, respective ones of which correspond to respective ones of the first plurality of user queries and that exhibit a first group of implicit user behavior with respect to the first plurality of user result sets;

receiving, at the first publication system, from the second publication system, a second plurality of user queries that are in a second language, that are in the category and that exhibit the constraint, and a second plurality of result sets, respective ones of which correspond to respective ones of the second plurality of user queries and that exhibit a second group of implicit user behavior with respect to the second plurality of result sets;

detecting, by a user behavior module of a machine translation application in the second publication system, implicit user behavior in the first group of implicit user behavior that indicates that at least some of the first plurality of user queries are satisfied by respective ones of the first plurality of result sets;

detecting, by a user behavior module of a machine translation application in the first publication system, implicit user behavior in the second group of implicit user behavior that indicates that at least some of the second plurality of user queries are satisfied by respective ones of the second plurality of result sets;

detecting that the at least some of the first plurality of user queries and the at least some of the second plurality of user queries satisfy respective thresholds; and

responsive to detecting that the at least some of the first plurality of first user queries and the at least some of the second plurality of user queries satisfy respective thresholds, providing an indication to at least one of the machine translation application in the second publication system or the machine translation application in the first publication system that at least one of the first plurality of user queries and at least one of the second plurality of user queries are likely to be translations of each other.

6. The one or more computer-readable hardware storage device of claim 5 wherein the constraint comprises one of aspects, a bar code or a seller.

7. The one or more computer-readable hardware storage device of claim 5 wherein the first group of implicit user behavior and the second group of implicit user behavior comprise one of positive behavior, potentially negative behavior or clearly negative behavior.

8. The one or more computer-readable hardware storage device of claim 5 wherein the respective thresholds comprise a first percentage of the plurality of the first user queries that are satisfied by result sets responsive to the plurality of first user queries and a second percentage of the second plurality of user queries that are satisfied by result sets responsive to the plurality of second user queries, and the satisfaction is based on one of positive user behavior, potentially negative user behavior, or clearly negative user behavior.

9. A system comprising:

one or more computer processors and storage, at a second publication system, configured to;

receive, from a first publication system, a first plurality of user queries that are in a first language that are in a category and that exhibit a constraint, and a first plurality of result sets, respective ones of which correspond to respective ones of the first plurality of user queries and that exhibit a first group of implicit user behavior with respect to the first plurality of user result sets;

detect, by a user behavior module of a machine translation application at the second publication system, implicit user behavior in the first group of implicit user behavior that indicates that at least some of the first plurality of user queries are satisfied by respective ones of the first plurality result sets;

one or more computer processors and storage, at the first publication system, configured to;

receive, from the second publication system, a second plurality of user queries in a second language that that are in the category and that exhibit the constraint, and a second plurality of result sets that exhibit a second group of implicit user behavior with respect to the second plurality of result sets;

detect by a user behavior module of a machine translation application at the first publication system, implicit user behavior in the second group of implicit user behavior that indicates that at least some of the second plurality of user queries are satisfied by respective ones of the second plurality of result sets;

execute a decision application configured to detect that the at least some of the first plurality of user queries and the at least some of the second plurality of user queries satisfy respective thresholds; and

execute a learning and correlation module configured, responsive to the detection that the at least some of the first plurality of first user queries and the at least some of the second plurality of user queries satisfy respective thresholds, to provide an indication to at least one of the machine translation application in the second publication system or the machine translation application in the first publication system that at least one of the first plurality of user queries and at least one of the second plurality of user queries are likely to be translations of each other.

10. The system of claim 9 wherein the constraint comprises one of aspects, a bar code or a seller.

11. The system of claim 9 wherein the first group of implicit user behavior and the second group of implicit user behavior comprises one of positive behavior, potentially negative behavior or clearly negative behavior.

12. The system of claim 9 wherein the respective thresholds comprise a first percentage of the plurality of the first user queries that are satisfied by result sets responsive to the plurality of first user queries and a second percentage of the second plurality of user queries that are satisfied by result sets responsive to the second plurality of user queries.