Joint ranking model for multilingual web search
First Claim
1. In a computing environment, a method comprising:
- determining similarity between a first document of a first language and a second document of a second, different language;
ranking, by a Boltzmann machine, relevance of the second document with respect to a query submitted in the first language using the similarity determined; and
displaying a result set according to the relevance in response to the query submitted in the first language, wherein the result set includes at least the second document of the second, different language.
2 Assignments
0 Petitions
Accused Products
Abstract
A classifier is built to rank documents of different languages found in a query based at least in part on similarity to other documents and the relevance of those other documents to the query. A joint ranking model, e.g., based upon a Boltzmann machine, is used to represent the content similarity among documents, and to help determine joint relevance probability for a set of documents. The relevant documents of one language are thus leveraged to improve the relevance estimation for documents of different languages. In one aspect, a hidden layer of units (neurons) represents clusters (corresponding to relevant topics) among the retrieved documents, with an output layer representing the relevant documents and their features, and edges representing a relationship between clusters and documents.
19 Citations
20 Claims
-
1. In a computing environment, a method comprising:
-
determining similarity between a first document of a first language and a second document of a second, different language; ranking, by a Boltzmann machine, relevance of the second document with respect to a query submitted in the first language using the similarity determined; and displaying a result set according to the relevance in response to the query submitted in the first language, wherein the result set includes at least the second document of the second, different language. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9)
-
-
10. In a computing environment, a system comprising:
-
a Boltzmann machine configured to rank multilingual documents relative to one another, including relating features of documents to topic-based clusters of documents, the Boltzmann machine further configured to rank documents of different languages relative to a query based upon similarity data between documents according to features of those documents; and a display configured to display a set of documents in response to the query, wherein the display of the set of documents is ordered at least in part based on results from the Boltzmann machine. - View Dependent Claims (11, 12, 13, 14, 15)
-
-
16. One or more computer storage media having computer-executable instructions stored thereon, which in response to execution by a computer, cause the computer to perform steps comprising:
-
featurizing a set of documents of different languages into multilingual features; clustering the set of documents into topic-based clustering data using the multilingual features; constructing a Boltzmann machine based on the documents, the multilingual features of the documents, and the topic-based clustering data; using the Boltzmann machine to rank the set of documents of different languages returned in a search; and displaying at least a portion of the set of documents, wherein an order to display the portion of the set of documents is based at least in part on the rank generated by the Boltzmann machine. - View Dependent Claims (17, 18, 19, 20)
-
Specification