Method and apparatus for recommendation engine using pair-wise co-occurrence consistency

A method for implementation by one or more data processors comprising:

• receiving transaction data comprising a plurality of purchase transactions of a plurality of products by a plurality of consumers;
• identifying patterns of interest in pair-wise relationships in the transaction data between products and purchasers using a pair-wise co-occurrence consistency framework, the pair-wise co-occurrence consistency framework comprising a semi-supervised insight discovery and data-drive decision analytics framework, the pair-wise relationship being represented in a graph structure containing a set of nodes representing entities, a set of edges representing strength of relationships between pairs of nodes, a weighted edge being used between each pair of nodes to represent a consistency with which products in particular categories are purchased together, edges with weights below a predetermined threshold being ignored;
• and
• initiating one or more subsequent transactions having offerings optimized using the identified patterns of interest.

A method for implementation by one or more data processors, the method comprising:

• applying, by at least one data processor, consistency and similarity functions to transaction data, the transaction data generated in connection with a plurality of product purchases by a plurality of customers involving a plurality of products;
• determining, by at least one data processor, statistically significant and logical associations between products based on the applying of consistency and similarity functions to the transaction data;
• analyzing, by at least one data processor, product associations in a plurality of contexts based on the determination of the statistically significant and logical associations;
• initiating by at least one data processor, one or more subsequent transactions based on the analyzed product associations;
• and
• representing, by at least one data processor, pair-wise relationships between entities abstraction in a graph structure containing a set of nodes representing entities, and a set of edges representing strength of relationships between pairs of nodes, a weighted edge being used between each pair of nodes to represent a consistency with which products in particular categories are purchased together, edges with weights below a predetermined threshold being ignored.

A method for implementation by one or more data processors 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 processors, pair-wise relationships between entities in the selected transaction data slice based on a context parameter and using a air-wise co-occurrence consistency framework that comprises a semi-supervised insight discovery and data-drive decision analytics framework, each pair-wise relationship linking two entities to each other;
• and
• determining, by at least one data processors, a strength value of each pair-wise relationship, the strength value corresponding to a consistency parameter;
• transforming, by at least one data processors, 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 processors, 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.