Query revision using known highly-ranked queries

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.

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.

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.

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.

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.

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.