Item substitution techniques for assortment optimization and product fulfillment

Patent metadata

Field	Value
Publication number	US-2021166179-A1
Application number	US-202117174667-A
Country	US
Kind code	A1
Filing date	Feb 12, 2021
Priority date	Feb 1, 2019
Publication date	Jun 3, 2021
Grant date	—

Abstract

Methods and systems for optimizing a product assortment, and managing product fulfillment, are disclosed. One method includes utilizing a model trained on item data to identify item substitution pairs within an item category of an item assortment, the item substitution pairs being identified as having a substitutability score above a predetermined threshold. The method further includes applying an assortment optimization model to generate an assortment recommendation for the item category at an identified retail location. A request for an item not included in the assortment recommendation may result in suggestion of an identified substitutable item from among the item substitution pairs.

First claim

1 . A method of determining an item assortment comprising:
utilizing a model trained on item data and item selection data to identify one or more item substitution pairs within an item category of an item assortment, the item substitution pairs being identified as having a substitutability score above a predetermined threshold; applying an assortment optimization model to generate an assortment recommendation for the item category at an identified retail location based at least in part on a set of items within the item category available at the identified retail location, a sales forecast for items in the set of items, a customer loyalty factor, the one or more item substitution pairs, and one or more physical constraints specific to the retail location.

2 . The method of claim 1 , wherein the item selection data includes item purchase data derived from online purchase transactions.

3 . The method of claim 2 , further comprising applying item-to-item collaborative filtering to each of the plurality of pairs of items to identify one or more substitutable item pairs within the plurality of pairs of items.

4 . The method of claim 1 , wherein the model comprises a graph convolutional network utilizing weighted graphs.

5 . The method of claim 4 , wherein the graph convolutional network is trained based at least in part on online guest transactions.

6 . The method of claim 5 , wherein the graph convolutional network is trained based at least in part on substitution subgraphs generated from portions of the item assortment.

7 . The method of claim 1 , wherein the assortment recommendation is further based in part on one or more business rules including a rule defining items within the item category indicated as being mandatory to be stocked at the identified retail location.

8 . The method of claim 1 , wherein the assortment optimization model maximizes sales volume of items across the item category subject to constraints of (1) a maximum number of items selected for stocking within the item category, (2) a minimum number of items selected for stocking within the item category, and (3) a business requirement of an item having an item loyalty greater than a predetermined threshold.

9 . The method of claim 8 , wherein the assortment optimization model includes at least one substitute item from a substitutable item pair in place of another item within the substitutable item pair based at least in part on one or more of the constraints.

10 . The method of claim 8 , wherein the assortment optimization model is:

max
i



=

∑

i
=
1

N





x
i



s
i

+

∑

i
=
1

N





∑

j
=
1

N





s
i



p
ij



x
j



(

1
-

x
i

)

wherein i is the item at issue in an item category, from among items 1 to N, x is a binary indicating whether an item is included in an assortment, s is a forecasted sales volume, j represents a substitute of item i, and p represents a proportion of customers who would substitute item i for item j, subject to constraints:

∑

i
=
1

N





x
i

≤
MaxItemCount

∑

i
=
1

N





x
i

≥
MinItemCount

x
i

≥

loyalty
i

-

loyalty
threshold

.

11 . The method of claim 1 , further comprising:
receiving a request for an item in the item assortment that is excluded from the assortment recommendation; and presenting, in a user interface, a recommendation of an item within the assortment recommendation that is in stock at the retail location.

12 . The method of claim 11 , wherein the request for the item comprises a request for same-day delivery of the item from the identified retail location.

13 . A system comprising:
a computing system comprising one or more processors communicatively coupled to a memory subsystem that stores instructions which, when executed, cause the one or more

Assignees

• Target Brands Inc

Inventors

• Pande Amit
• Das Gupta Aparupa
• Ni Kai
• Biswas Rahul
• Majumdar Sayon

Classifications

• G06N3/045

  Combinations of networks · CPC title

• G06N3/047

  Probabilistic or stochastic networks · CPC title

• G06N3/044

  Recurrent networks, e.g. Hopfield networks · CPC title

• G06N3/0895

  Weakly supervised learning, e.g. semi-supervised or self-supervised learning · CPC title

• G06N3/0464

  Convolutional networks [CNN, ConvNet] · CPC title