Retrieval of relevant objects in a similarity

A computer program product comprising a non-transitory tangible computer readable recordable storage medium including computer useable program code for retrieval of one or more relevant multi-attribute structured objects with respect to a query, the computer program product including:

• computer useable program code for receiving a user query for retrieval of one or more multi-attribute structured objects from a database that includes multiple objects with interdependent and independent attributes;
• computer useable program code for receiving user instructions specifying how the interdependent and independent attributes are to be grouped in carrying out the query;
• computer useable program code for grouping the attributes of the objects in the database into one or more groups according to the user instructions by:
• associating each of the attributes with a list of one or more database objects that are ordered in an increasing value of dissimilarity from a query value for the respective attribute;
• combining multiple aggregation operators in accordance with user-provided specifications, wherein the multiple aggregation operators comprise:
• (i) an intersection aggregation operator, wherein the intersection aggregation operator deems a first object as having a lower value of dissimilarity than a second object if the most dissimilar attribute of the first object has a higher absolute value than the most dissimilar attribute of the second object,
• (ii) a union aggregation operator, wherein the union aggregation operator deems a first object as having a lower value of dissimilarity than a second object if the least dissimilar attribute of the first object has a higher absolute value than the least dissimilar attribute of the second object,
• (iii) a generalized mean aggregation operator, and
• (iv) a Euclidean distance aggregation operator, to compose the interdependent attributes into multiple interdependent attribute compositions;
• and
• using a skyline operator to combine the multiple interdependent attribute compositions and the independent attributes into the one or more groups according to the user instructions;
• and
• computer useable program code for using the one or more attribute groups to produce an output of a sorted list of one or more relevant multi-attribute structured objects from the database in response to the query that express preference specification of the one or more attributes.

A system for retrieval of one or more relevant multi-attribute structured objects with respect to a query, comprising:

• a memory;
• and
• at least one processor coupled to the memory and operative to:
• receive a user query for retrieval of one or more multi-attribute structured objects from a database that includes multiple objects with interdependent and independent attributes;
• receive user instructions specifying how the interdependent and independent attributes are to be grouped in carrying out the query;
• group the attributes of the objects in the database into one or more groups according to the user instructions by:
• associating each of the attributes with a list of one or more database objects that are ordered in an increasing value of dissimilarity from a query value for the respective attribute;
• combining multiple aggregation operators in accordance with user-provided specifications, wherein the multiple aggregation operators comprise:
• (i) an intersection aggregation operator, wherein the intersection aggregation operator deems a first object as having a lower value of dissimilarity than a second object if the most dissimilar attribute of the first object has a higher absolute value than the most dissimilar attribute of the second object,
• (ii) a union aggregation operator, wherein the union aggregation operator deems a first object as having a lower value of dissimilarity than a second object if the least dissimilar attribute of the first object has a higher absolute value than the least dissimilar attribute of the second object,
• (iii) a generalized mean aggregation operator, and
• (iv) a Euclidean distance aggregation operator, to compose the interdependent attributes into multiple interdependent attribute compositions;
• and
• using a skyline operator to combine the multiple interdependent attribute compositions and the independent attributes into the one or more groups according to the user instructions;
• and
• use the one or more attribute groups to produce an output of a sorted list of one or more relevant multi-attribute structured objects from the database in response to the query that express preference specification of the one or more attributes.