Method and apparatus for recommendation engine using pair-wise co-occurrence consistency
First Claim
1. An article comprising a tangible machine-readable storage medium embodying instructions that when performed by one or more processors result in operations comprising:
- selecting a transaction data slice based on a customization parameter, the transactional data slice corresponding to a plurality of entities, the transaction data slice characterizing a time stamped sequence of market baskets, each market basket characterizing a transaction of a set of one or more products;
generating pair-wise relationships between entities in the selected transaction data slice based on a context parameter, each pair-wise relationship linking two entities to each other;
determining a strength value of each pair-wise relationship, the strength value corresponding to a consistency parameter;
transforming data corresponding to at least one pair-wise relationship and at least one strength value into a recommendation; and
initiating a visual presentation of the recommendation in the form of a graph comprising a plurality of nodes and edges, the nodes corresponding to entities and the edges corresponding to strength values, a structure of the graph being selected from a group comprising;
(a) a sub-graph comprising a subset of a graph, created by picking a subset of nodes and edges from an original graph, the sub-graph selected from a group comprising (i) node based sub-graphs which are created by selecting a subset of the nodes and by keeping only those edges between selected nodes, and (ii) edge based sub-graphs which are created by pruning a set of edges from the graph and removing all nodes that are rendered disconnected from the graph;
(b) a neighborhood of a target product comprising a sub-graph that contains the target product and all the products that are connected to the target product with consistency strength above a predefined threshold to show the top most affiliated products for a given target product;
(c) a bundle structure comprising a sub-set of products wherein each product in the bundle has a high consistency connection with all the other products in the bundle, wherein each product in a bundle is assigned a product density with respect to the bundle which is high if the product has high consistency connection with other products in the bundle and low otherwise; and
(d) a bridge structure comprising a collection of two or more, otherwise disconnected, product groups that are bridged by one or more bridge product.
1 Assignment
0 Petitions
Accused Products
Abstract
The invention, referred to herein as PeaCoCk, uses a unique blend of technologies from statistics, information theory, and graph theory to quantify and discover patterns in relationships between entities, such as products and customers, as evidenced by purchase behavior. In contrast to traditional purchase-frequency based market basket analysis techniques, such as association rules which mostly generate obvious and spurious associations, PeaCoCk employs information-theoretic notions of consistency and similarity, which allows robust statistical analysis of the true, statistically significant, and logical associations between products. Therefore, PeaCoCk lends itself to reliable, robust predictive analytics based on purchase-behavior.
-
Citations
34 Claims
-
1. An article comprising a tangible machine-readable storage medium embodying instructions that when performed by one or more processors result in operations comprising:
-
selecting a transaction data slice based on a customization parameter, the transactional data slice corresponding to a plurality of entities, the transaction data slice characterizing a time stamped sequence of market baskets, each market basket characterizing a transaction of a set of one or more products; generating pair-wise relationships between entities in the selected transaction data slice based on a context parameter, each pair-wise relationship linking two entities to each other; determining a strength value of each pair-wise relationship, the strength value corresponding to a consistency parameter; transforming data corresponding to at least one pair-wise relationship and at least one strength value into a recommendation; and initiating a visual presentation of the recommendation in the form of a graph comprising a plurality of nodes and edges, the nodes corresponding to entities and the edges corresponding to strength values, a structure of the graph being selected from a group comprising; (a) a sub-graph comprising a subset of a graph, created by picking a subset of nodes and edges from an original graph, the sub-graph selected from a group comprising (i) node based sub-graphs which are created by selecting a subset of the nodes and by keeping only those edges between selected nodes, and (ii) edge based sub-graphs which are created by pruning a set of edges from the graph and removing all nodes that are rendered disconnected from the graph; (b) a neighborhood of a target product comprising a sub-graph that contains the target product and all the products that are connected to the target product with consistency strength above a predefined threshold to show the top most affiliated products for a given target product; (c) a bundle structure comprising a sub-set of products wherein each product in the bundle has a high consistency connection with all the other products in the bundle, wherein each product in a bundle is assigned a product density with respect to the bundle which is high if the product has high consistency connection with other products in the bundle and low otherwise; and (d) a bridge structure comprising a collection of two or more, otherwise disconnected, product groups that are bridged by one or more bridge product. - 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, 28, 29, 30, 31, 32)
-
-
33. A method for implementation by one or more data processors, the method comprising:
-
selecting, by at least one data processor, a transaction data slice based on a customization parameter, the transactional data slice corresponding to a plurality of entities, the transaction data slice characterizing a time stamped sequence of market baskets, each market basket characterizing a transaction of a set of one or more products; generating, by at least one data processor, pair-wise relationships between entities in the selected transaction data slice based on a context parameter, each pair-wise relationship linking two entities to each other; determining, by at least one data processor, a strength value of each pair-wise relationship, the strength value corresponding to a consistency parameter; transforming, by at least one data processor, data corresponding to at least one pair-wise relationship and at least one strength value into a recommendation; and initiating, by at least one data processor, a visual presentation of the recommendation in the form of a graph comprising a plurality of nodes and edges, the nodes corresponding to entities and the edges corresponding to strength values, a structure of the graph being selected from a group comprising; a sub-graph comprising a subset of a graph, created by picking a subset of nodes and edges from an original graph, the sub-graph selected from a group comprising (i) node based sub-graphs which are created by selecting a subset of the nodes and by keeping only those edges between selected nodes, and (ii) edge based sub-graphs which are created by pruning a set of edges from the graph and removing all nodes that are rendered disconnected from the graph; a neighborhood of a target product comprising a sub-graph that contains the target product and all the products that are connected to the target product with consistency strength above a predefined threshold to show the top most affiliated products for a given target product; a bundle structure comprising a sub-set of products wherein each product in the bundle has a high consistency connection with all the other products in the bundle, wherein each product in a bundle is assigned a product density with respect to the bundle which is high if the product has high consistency connection with other products in the bundle and low otherwise; and a bridge structure comprising a collection of two or more, otherwise disconnected, product groups that are bridged by one or more bridge product.
-
-
34. A system comprising:
-
means for selecting a transaction data slice based on a customization parameter, the transactional data slice corresponding to a plurality of entities, the transaction data slice characterizing a time stamped sequence of market baskets, each market basket characterizing a transaction of a set of one or more products; means for generating pair-wise relationships between entities in the selected transaction data slice based on a context parameter, each pair-wise relationship linking two entities to each other; means for determining a strength value of each pair-wise relationship, the strength value corresponding to a consistency parameter; means for transforming data corresponding to at least one pair-wise relationship and at least one strength value into a recommendation; and means for initiating a visual presentation of the recommendation in the form of a graph comprising a plurality of nodes and edges, the nodes corresponding to entities and the edges corresponding to strength values, a structure of the graph being selected from a group comprising; a sub-graph comprising a subset of a graph, created by picking a subset of nodes and edges from an original graph, the sub-graph selected from a group comprising (i) node based sub-graphs which are created by selecting a subset of the nodes and by keeping only those edges between selected nodes, and (ii) edge based sub-graphs which are created by pruning a set of edges from the graph and removing all nodes that are rendered disconnected from the graph; a neighborhood of a target product comprising a sub-graph that contains the target product and all the products that are connected to the target product with consistency strength above a predefined threshold to show the top most affiliated products for a given target product; a bundle structure comprising a sub-set of products wherein each product in the bundle has a high consistency connection with all the other products in the bundle, wherein each product in a bundle is assigned a product density with respect to the bundle which is high if the product has high consistency connection with other products in the bundle and low otherwise; and a bridge structure comprising a collection of two or more, otherwise disconnected, product groups that are bridged by one or more bridge product.
-
Specification