Query revision using known highly-ranked queries
First Claim
1. A method for automatically suggesting known highly-ranked queries in response to a first query, comprising:
- calculating a revision score for an indexed query as a function of a revision probability for the first query with respect to the indexed query and a query rank for the indexed query;
responsive to the revision score, selectively retrieving the indexed query as an alternative query to the first query; and
responsive to the alternative query being a known highly-ranked query, returning the alternative query as a candidate revision query.
2 Assignments
0 Petitions
Accused Products
Abstract
An information retrieval system includes a query revision architecture providing one or more query revisers, each of which implements a query revision strategy. A query rank reviser suggests known highly-ranked queries as revisions to a first query by initially assigning a rank to all queries, and identifying a set of known highly-ranked queries (KHRQ). Queries with a strong probability of being revised to a KHRQ are identified as nearby queries (NQ). Alternative queries that are KHRQs are provided as candidate revisions for a given query. For alternative queries that are NQs, the corresponding known highly-ranked queries are provided as candidate revisions.
205 Citations
42 Claims
-
1. A method for automatically suggesting known highly-ranked queries in response to a first query, comprising:
-
calculating a revision score for an indexed query as a function of a revision probability for the first query with respect to the indexed query and a query rank for the indexed query;
responsive to the revision score, selectively retrieving the indexed query as an alternative query to the first query; and
responsive to the alternative query being a known highly-ranked query, returning the alternative query as a candidate revision query. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18)
-
-
19. A method for automatically suggesting known highly-ranked queries in response to a first query, comprising:
-
logging query data generated from user sessions;
creating an index of queries during the user session;
calculating a revision score for an indexed query as a function of a revision probability for the first query with respect to the indexed query and a query rank for the indexed query, wherein the revision probability comprises the similarity of the indexed query with respect to the first query;
responsive to the revision score, selectively retrieving the indexed query as an alternative query to the first query;
responsive to the alternative query being a known highly-ranked query, returning the alternative query as a candidate revision query, wherein identifying the known highly-ranked query comprises;
calculating a query occurrence frequency for a query;
calculating a user satisfaction score for the query, wherein the user satisfaction score is determined by user click behavior data estimating the length of clicks on search results; and
computing a rank for the query as a product of the query occurrence frequency and the user satisfaction score;
responsive to the alternative query having a statistically significant probability of revising to a known highly-ranked query, returning the known highly-ranked query as a candidate revision query;
ranking the candidate revision query using the revision score for the candidate revision query as a confidence measure; and
providing the candidate revision query as a suggested revision for the first query, wherein the suggested revision is displayed to a user in a location dependent upon a relative strength of the confidence measure.
-
-
20. A method of identifying a query as a known highly-ranked query, comprising:
-
calculating a query occurrence frequency for the query;
calculating a user satisfaction score for the query; and
computing a query rank for the query using the query occurrence frequency and the user satisfaction score. - View Dependent Claims (21, 22, 23, 24, 25, 26, 27, 28)
-
-
29. A computer program product for automatically suggesting known highly-ranked queries in response to a first query, the computer program product comprising:
-
a computer-readable medium; and
computer program code, coded on the medium, for;
calculating a revision score for an indexed query as a function of a revision probability for the first query with respect to the indexed query and a query rank for the indexed query;
responsive to the revision score, selectively retrieving the indexed query as an alternative query to the first query; and
responsive to the alternative query being a known highly-ranked query, returning the alternative query as a candidate revision query. - View Dependent Claims (30, 31)
-
-
32. A computer program product for identifying a query as a known highly-ranked query, the computer program product comprising:
-
a computer-readable medium; and
computer program code, coded on the medium, for;
calculating a query occurrence frequency for the query;
calculating a user satisfaction score for the query; and
computing a query rank using the query occurrence frequency and the user satisfaction score. - View Dependent Claims (33, 34, 35)
-
-
36. A system for providing revised queries for a query as a known highly-ranked query, the system comprising:
-
means for calculating a revision score for an indexed query as a function of a revision probability for the first query with respect to the indexed query and a query rank for the indexed query;
means for responsive to the revision score, selectively retrieving the indexed query as an alternative query to the first query; and
means for responsive to the alternative query being a known highly-ranked query, returning the alternative query as a candidate revision query. - View Dependent Claims (37, 38)
-
-
39. A system for providing revised queries for identifying a query as a known highly-ranked query, the system comprising:
-
means for calculating a query occurrence frequency for the query;
calculating a user satisfaction score for the query; and
computing a query rank using the query occurrence frequency and the user satisfaction score. - View Dependent Claims (40, 41, 42)
-
Specification