Back to results
US 8,620,891 B1Granted
Ranking item attribute refinements
AssigneeAmazon Technologies, Inc.
InventorSrikanth Thirumalai
Life of this patent
Estimating the term…
Abstract
Systems and methods for enabling refinement of information sets utilize recorded instances of user behavior in connection with information units of the information sets. Information units are each associated with one or more attributes and provided to users in ranked information sets. Instances of user selection of information units are used to determine how refining search result sets according to attributes would have improved the positions of the selected information units. Information is accumulated accordingly to determine how to present representations of attributes to users to enable the users to refine information sets.
Claims25
A computer-implemented method for presenting search results, comprising:
- under the control of one or more computer systems configured with executable instructions,
- for each search result of a plurality of search results that have been selected by users from corresponding search result sets, identifying one or more features associated with the search result;
- for each identified feature of the one or more features, determining a potential improvement in position for at least one selected search result within a corresponding search result set, the potential improvement including a number of places that the at least one selected search result would have moved within the corresponding search result set had an identified feature of the one or more features been selected and achievable by refining the corresponding search result set according to the identified feature of the one or more features;
- using the determined potential improvement to calculate a feature score for the identified feature of the one or more features;
- for each of a plurality of search queries, ranking a set of features based at least in part on corresponding calculated feature scores for the identified features of the one or more features;
- and
- for at least one subsequently received search query from the plurality of search queries, providing a corresponding ranked set of features in connection with a search result set responsive to the subsequently received search query to enable user refinement of the search result set according to one or more features in the provided ranked set of features.
4 more elements
A computer-implemented method for enabling information set modification, comprising:
- under the control of one or more computer systems configured with executable instructions,
- for each information unit from a plurality of information units that has been selected from a corresponding information set, identifying one or more attributes associated with the information unit;
- for each identified attribute of the one or more attributes, determining a potential improvement in position for at least one user-selected information unit achievable by modifying a corresponding information set according to an identified attribute of the one or more attributes, the potential improvement including a number of places that the at least one user-selected information unit would have moved within the corresponding information set had the identified attribute of the one or more attributes been selected;
- ranking at least one attribute set based at least in part on the determined potential improvements;
- and
- providing the ranked attribute set to enable a user to modify an information set provided to the user.
3 more elements
A computer system for enabling information set modification, comprising:
- one or more processors;
- and
- memory, including instructions executable by the one or more processors to cause the computer system to at least:
- calculate scores for attributes based at least in part on measurements of potential position changes within information sets for information units that users have selected and that are each associated with at least one of the attributes, the potential position changes including a number of position places the information units that the users have selected would have moved within the information sets had the at least one of the attributes been selected;
- for each of a plurality of queries, rank a set of attributes based at least in part on the calculated attribute scores for the information sets;
- and
- in response to receipt of a query from the plurality of queries, provide a corresponding ranked set of attributes in connection with an information sets responsive to the received query to enable user modification of the information set responsive to the received query according to one or more attributes in the provided ranked set of attributes.
4 more elements
One or more non-transitory computer-readable storage media having collectively stored thereon executable instructions that, when executed by one or more processors of a computer system, cause the computer system to at least:
- for each information unit of multiple information units that has been selected by a user from a ranked information set:
- calculate a rank improvement score for at least one attribute associated with the information unit, the rank improvement score being based at least in part on a potential ranking improvement for the information unit that is achievable for the information unit by modifying the information set according to the at least one attribute associated with the information unit, the potential ranking improvement including a number of places that the information unit would have moved within the information set had the at least one attribute associated with the information unit been selected;
- and
- update a cumulative rank improvement score for the at least one attribute associated with the information unit based at least in part on the rank improvement score, the cumulative rank improvement score being based at least in part on previously calculated rank improvement scores for the at least one attribute associated with the information unit;
- and
- for each query of a plurality of queries, provide a ranking of attributes based at least in part on corresponding cumulative rank improvement scores to enable modification of information sets according to the ranked attributes.
3 more elements
Dependencies are inferred from the claim text.