Similarity matching of products based on multiple classification schemes
View Patent ↗In embodiments of the present invention improved capabilities are described for identifying a first classification scheme associated with product attributes of a first grouping of products, identifying a second classification scheme associated with product attributes of a second grouping of products, and receiving a record of data relating to an item, the classification of which is uncertain. It may also involve receiving a dictionary of attributes associated with products and assigning the item to at least one of the classification schemes based on probabilistic matching among the attributes in the classification schemes, the attributes in the dictionary of attributes and the known attributes of the item.
1. A method for adding an unclassified item to a hierarchical arrangement of products, the method comprising:
identifying a first hierarchy of products, the first hierarchy providing a first classification scheme associated with product attributes of a first grouping of products in a first category, wherein each product in the first hierarchy is included in a block of the first hierarchy having known block attributes, and each product in the first hierarchy has known match attributes;
identifying a second hierarchy of products, the second hierarchy providing a second classification scheme associated with product attributes of a second grouping of products in a second category that classifies a different type of product than the first hierarchy of products, wherein each product is included in a block of the second hierarchy having known block attributes, and each product in the second hierarchy has known match attributes;
using a computer, receiving a record of attributes of an unclassified item;
assigning the item to a location in at least one of the first or second hierarchies based on probabilistic matching among the attributes of products in the hierarchies, and the known attributes of the item, thereby classifying the item in a location in the first hierarchy or the second hierarchy; and
after assigning the item to the location, generating new attribute data based on a product type corresponding to the location of the item and existing attribute data and associating the new attribute data with the item.
2. The method of claim 1 , further comprising iterating the probabilistic matching until a statistical criterion is met.
3. The method of claim 1 , wherein the known attributes of the item include a nutritional level.
4. The method of claim 1 , wherein the attributes of the item include a brand.
5. The method of claim 1 , wherein the attributes of the item include product category.
6. The method of claim 1 , wherein the attributes of the item include an attribute based at least in part on a SKU.
7. The method of claim 1 , wherein the attributes of the item include physical attribute.
8. The method of claim 7 , wherein the physical attribute is a flavor.
9. The method of claim 7 , wherein the physical attribute is a scent.
10. The method of claim 7 , wherein the physical attribute is packaging type.
11. The method of claim 7 , wherein the physical attribute is a product launch date.
12. The method of claim 7 , wherein the physical attribute is display location.
13. The method of claim 1 , wherein assigning the item to one of the hierarchies includes assigning the item to one of the hierarchies based upon probabilistic matching to the attributes in a dictionary of attributes for one of the hierarchies.