Listings optimization using a plurality of data sources
First Claim
1. A method comprising:
- receiving a query;
receiving a plurality of items from a plurality of candidate sources;
generating an initial set of possible composite search results, each of which is a different composition of items from the plurality of candidate sources;
determining a first estimated attribute value and a second estimated attribute value for each possible composite search result in the initial set of possible composite search results;
filtering the initial set of possible composite search results based upon whether the first estimated attribute value satisfies a specified threshold value to generate a filtered set;
selecting a final composite search result from the filtered set where the second estimated attribute value is greatest; and
providing for display at a client the final composite search result as the answer to said query,wherein one of the estimated attribute values is a commercial value and the other estimated attribute value is a user satisfaction value, andwherein the method is performed by one or more computing devices.
9 Assignments
0 Petitions
Accused Products
Abstract
A system is provided for allocating combinations of search results, types of space, and/or other information from different sources (e.g., search engines) in which a type of space, search results, and/or other information from at least one source or type or from each source or type, for example, are optimized to obtain an optimum commercial value. Using the example of a search, a commercial metric and a user satisfaction metric are established for search terms and used in the optimization process. In calculating the commercial value and the user satisfaction value for a set of search results, the weight given to the commercial value and the user satisfaction value of a search result is different depending on its source. Optionally, keeping the user satisfaction value above a certain threshold is treated as a constraint while optimizing the commercial value. Alternatively, a tradeoff may be allowed between the minimum threshold for the user satisfaction and a better commercial value. For example, a commercial value may be associated with the commercial metric and quality metric so that the total commercial value may be calculated and optimized.
-
Citations
32 Claims
-
1. A method comprising:
-
receiving a query; receiving a plurality of items from a plurality of candidate sources; generating an initial set of possible composite search results, each of which is a different composition of items from the plurality of candidate sources; determining a first estimated attribute value and a second estimated attribute value for each possible composite search result in the initial set of possible composite search results; filtering the initial set of possible composite search results based upon whether the first estimated attribute value satisfies a specified threshold value to generate a filtered set; selecting a final composite search result from the filtered set where the second estimated attribute value is greatest; and providing for display at a client the final composite search result as the answer to said query, wherein one of the estimated attribute values is a commercial value and the other estimated attribute value is a user satisfaction value, and wherein the method is performed by one or more computing devices. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26)
-
-
27. A method comprising:
-
receiving a query; receiving a plurality of items from a plurality of candidate sources; generating a plurality of item groups, each of which is a different composition of items from the plurality of candidate sources; determining a first estimated attribute value and a second estimated attribute value for each item group in the plurality of item groups; selecting a final item group from the plurality of item sets based upon first estimated attribute values and second estimated attribute values for each item group in the plurality of item groups; and providing, for display at a client, the final item group as an answer to said query, wherein one of the estimated attribute values is a commercial value and the other estimated attribute value is a quality value, and wherein the method is performed by one or more computing devices. - View Dependent Claims (28, 29, 30)
-
-
31. A method comprising:
-
receiving a query; receiving, based upon the query, a plurality of items from a plurality of candidate sources; generating an initial set of item groups, where each item group in the initial set of item groups is a different composition of items from the plurality of candidate sources; determining a commercial value, that is a measure of the commercial nature of the query, and a quality value, that is an expected measure of customer satisfaction with the items, for each item group in the initial set of item groups; filtering the initial set of item groups based upon whether the quality value is greater than a specified threshold value to generate a filtered set; selecting a final item group from the filtered set where the commercial value is greatest; and providing, for display at a client, the final item group as an answer to said query, wherein the method is performed by one or more computing devices. - View Dependent Claims (32)
-
Specification