# US 8,571,908 B2 · Allocating commodity shelves in a supermarket

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 8,571,908 B2
- Document ID: US8571908B2
- Status: Granted
- Kind: B2
- Application number: 13/336,845
- Filed: 2011-12-23
- Published: 2012-07-05
- Granted: 2013-10-29
- Earliest claimed: 2010-12-30
- Assignees: International Business Machines Corporation
- Applicants: Hong Bo Li; Wei Wang; Hong Wei Ding; Jin Dong
- Inventors: Hong Bo Li; Wei Wang; Hong Wei Ding; Jin Dong
- Web page: https://patentgrep.com/patent/US8571908B2

## CPC

- G06Q 30/0202: Market predictions or forecasting for commercial activities
- G06Q 30/02: Marketing; Price estimation or determination; Fundraising
- G06Q: INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR

## Abstract

A method of and a system for allocating commodity shelves including obtaining shopping paths of customers, classifying the customers according to a predetermined standard, determining one or more shopping paths adopted by more customers in each kind of customers as frequent shopping path(s) of this kind of customers, calculating a see-buy rate of a commodity for each class of customers. The see-buy rate of a commodity for a class of customers refers to a probability to purchase the commodity for the class of customers. Calculating a shelf where each commodity in a set of commodities is located when the total expected benefits for set of commodities are maximized during a certain period of time. The total expected benefits include a sum of an expected benefit for the each commodity based on the see-buy rate and the frequency shopping path when each commodity is located in its own shelf.

## Claims

### Claim 1

A method of allocating commodity shelves in a supermarket, the method comprising:
obtaining shopping paths of a plurality of customers in the supermarket;
classifying the plurality of customers, according to a predetermined standard, into a plurality of customer classes;
based on the shopping paths of the plurality of customers, determining, by a processor on a computer, one or more shopping paths adopted by more customers in the plurality of customer classes as frequent shopping paths of a class of customers of the plurality of customer classes;
calculating a see-buy rate of a commodity for each of the plurality of customer classes based on shopping lists and the frequent shopping paths of the plurality of customers, wherein the see-buy rate of the commodity for the class of customers refers to a probability to purchase the commodity for the class of customers when the class of customers sees the commodity;
calculating a location for a set of commodities for a shelf where each commodity in the set of commodities is located when total expected benefits for the set of commodities are maximized during a certain period of time,
wherein the total expected benefits include a sum of an expected benefit for each commodity in the set of commodities based on the see-buy rate and the frequent shopping paths when each commodity in the set of commodities is located in its own shelf, and
wherein, when placing an ith commodity on a jth shelf, total expected benefits P (i, j) which is based on the see-buy rate and the frequent shopping path(s) is:
P
⁡
(
i
,
j
)
=
profit
⁡
(
i
)
·
∑
k
⁢
[
num
⁡
(
k
,
j
)
·
R
bs
⁡
(
i
,
k
)
wherein k denotes the class of customers, Rbs(i, k) denotes a see-buy rate of the ith commodity for the kth class of customers, and num (k,j) denotes a number of actual shopping paths corresponding to the frequent shopping path(s) of the kth class of customers adjacent to the shelf j within the certain period of time, and profit (i) denotes a unit profit of the commodity i.

### Claim 2 (depends on 1)

The method according to claim 1, wherein the sum of the expected benefits for each commodity in the set of commodities is:
⁢
∑
i
,
j
⁢
P
⁡
(
i
,
j
)
·
x
i
,
j
⁢
wherein
,
⁢
x
ij
=
{
1
,
for
⁢
⁢
the
⁢
⁢
i
th
⁢
⁢
commodity
⁢
⁢
being
⁢
⁢
placed
⁢
⁢
on
⁢
⁢
the
⁢
⁢
j
th
⁢
⁢
shelf
0
,
for
⁢
⁢
the
⁢
⁢
i
th
⁢
⁢
commodity
⁢
⁢
not
⁢
⁢
being
⁢
⁢
placed
⁢
⁢
on
⁢
⁢
the
⁢
⁢
j
th
⁢
⁢
shelf
.

### Claim 3 (depends on 1)

The method according to claim 1, wherein the total expected benefits of the set of commodities further comprise a sum of associated benefits between a plurality of commodities in the set of commodities.

### Claim 4 (depends on 3)

The method according to claim 3, wherein the sum of associated benefits between the plurality of commodities in the set of commodities is:
∑
i
,
i
′
,
j
,
j
′
,
k
⁢
α
i
,
i
′
,
k
·
x
i
,
j
·
x
i
′
,
j
′
·
ϕ
⁡
(
j
,
j
′
,
k
)
wherein, αi,i′, k denotes associated benefits between the ith commodity and the i′th commodity for the kth class of customers, and
ϕ
⁡
(
j
,
j
′
,
k
)
=
{
1
,
for
⁢
⁢
the
⁢
⁢
shelf
⁢
⁢
j
th
⁢
⁢
and
⁢
⁢
the
⁢
⁢
shelf
⁢
⁢
j
′
⁢
⁢
th
⁢
⁢
being
located
⁢
⁢
simultaneously
⁢
⁢
along
the
⁢
⁢
frequent
⁢
⁢
path
⁢
⁢
of
⁢
⁢
the
⁢
⁢
k
th
⁢
⁢
class
⁢
⁢
of
⁢
⁢
customers
0
,
for
⁢
⁢
the
⁢
⁢
shelf
⁢
⁢
j
th
⁢
⁢
and
⁢
⁢
the
⁢
⁢
shelf
⁢
⁢
j
′
⁢
⁢
th
⁢
⁢
not
⁢
⁢
being
located
⁢
⁢
simultaneously
⁢
⁢
along
⁢
⁢
the
frequent
⁢
⁢
path
⁢
⁢
of
⁢
⁢
the
⁢
⁢
k
th
⁢
⁢
class
⁢
⁢
of
⁢
⁢
customers
⁢
⁢
x
ij
=
{
1
,
for
⁢
⁢
the
⁢
⁢
i
th
⁢
⁢
commodity
⁢
⁢
being
⁢
⁢
placed
⁢
⁢
on
⁢
⁢
the
⁢
⁢
j
th
⁢
⁢
shelf
0
,
for
⁢
⁢
the
⁢
⁢
i
th
⁢
⁢
commodity
⁢
⁢
not
⁢
⁢
being
⁢
⁢
placed
⁢
⁢
on
⁢
⁢
the
⁢
⁢
j
th
⁢
⁢
shelf
.

### Claim 5 (depends on 1)

The method according to claim 1, wherein said obtaining shopping paths of a plurality of customers in a supermarket comprises:
detecting an RFID tag on a shopping cart used by a customer of the plurality of customers through RFID readers deployed at different locations in the supermarket; and
recording dynamic locations of the shopping cart, and thereby deriving the shopping path of the customer.

### Claim 6 (depends on 1)

The method according to claim 1, wherein said classifying of the plurality of customers according to the predetermined standard comprises:
classifying the plurality of customers based on one or more of the following data sources:
shopping lists recorded by point of sale(POS) machines; and
demographic information of the plurality of customers.

### Claim 7

A system for allocating commodity shelves in a supermarket, comprising:
a shopping path obtaining device, configured to obtain shopping paths of customers in a supermarket;
a classifying device, configured to classify the customers according to a predetermined standard, into a plurality of customer classes;
a frequent shopping path determining device, comprising a processor on a computer, configured to, based on the shopping paths of the customers, determine one or more shopping paths adopted by more customers in each class of customers of the plurality of customer classes as frequent shopping paths of a class of customers of the plurality of customer classes;
a see-buy rate calculating device configured to calculate a see-buy rate of a commodity for each class of customers based on shopping lists and the shopping paths of the customers, wherein the see-buy rate of the commodity for the class of customers refers to a probability to purchase the commodity when it is seen by the class of customers;
a commodity shelf determining device, configured to, for a set of commodities, calculate a location for a shelf where each commodity in the set of commodities is located when total expected benefits for the set of commodities are maximized during a certain period of time,
wherein the total expected benefits include a sum of an expected benefit for each commodity of the plurality of commodities based on the see-buy rate and the frequent shopping path(s) when each commodity of the plurality of commodities is located in its own shelf, and
when placing an ith commodity on a jth shelf, an expected benefits P (i, j) based on the see-buy rate and the frequent shopping path(s) are:
P
⁡
(
i
,
j
)
=
profit
⁡
(
i
)
·
∑
k
⁢
[
num
⁡
(
k
,
j
)
·
R
bs
⁡
(
i
,
k
)
wherein k denotes the class of customers, Rbs(i, k) denotes a see-buy rate of the ith commodity for the kth class of customers, and num(k,j) denotes numbers of actual shopping paths corresponding to the frequent shopping path(s) of the kth class of customers adjacent to shelf j within a certain period of time, and profit (i) denotes a unit profit of the commodity i.

### Claim 8 (depends on 7)

The system according to claim 7, wherein the sum of the expected benefits for each commodity in the set of commodities is:
⁢
∑
i
,
j
⁢
P
⁡
(
i
,
j
)
·
x
i
,
j
⁢
wherein
,
⁢
x
ij
=
{
1
,
for
⁢
⁢
the
⁢
⁢
i
th
⁢
⁢
commodity
⁢
⁢
being
⁢
⁢
placed
⁢
⁢
on
⁢
⁢
the
⁢
⁢
j
th
⁢
⁢
shelf
0
,
for
⁢
⁢
the
⁢
⁢
i
th
⁢
⁢
commodity
⁢
⁢
not
⁢
⁢
being
⁢
⁢
placed
⁢
⁢
on
⁢
⁢
the
⁢
⁢
j
th
⁢
⁢
shelf
.

### Claim 9 (depends on 7)

The system according to claim 7, wherein the total expected benefits of the set of commodities further comprise a sum of associated benefits between a plurality of commodities in the set of commodities.

### Claim 10 (depends on 9)

The system according to claim 9, wherein the sum of associated benefits between the plurality of commodities in the set of commodities is:
∑
i
,
i
′
,
j
,
j
′
,
k
⁢
α
i
,
i
′
,
k
·
x
i
,
j
·
x
i
′
,
j
′
·
ϕ
⁡
(
j
,
j
′
,
k
)
wherein, αi,i′,k denotes associated benefits between the ith commodity and the i′th commodity for the kth class of customers, and
ϕ
⁡
(
j
,
j
′
,
k
)
=
{
1
,
for
⁢
⁢
the
⁢
⁢
shelf
⁢
⁢
j
th
⁢
⁢
and
⁢
⁢
the
⁢
⁢
shelf
⁢
⁢
j
′
⁢
⁢
th
⁢
⁢
being
located
⁢
⁢
simultaneously
⁢
⁢
along
the
⁢
⁢
frequent
⁢
⁢
path
⁢
⁢
of
⁢
⁢
the
⁢
⁢
k
th
⁢
⁢
class
⁢
⁢
of
⁢
⁢
customers
0
,
for
⁢
⁢
the
⁢
⁢
shelf
⁢
⁢
j
th
⁢
⁢
and
⁢
⁢
t
⁢
he
⁢
⁢
shelf
⁢
⁢
j
′
⁢
⁢
th
⁢
⁢
not
⁢
⁢
being
located
⁢
⁢
simultaneously
⁢
⁢
along
the
⁢
⁢
frequent
⁢
⁢
path
⁢
⁢
of
⁢
⁢
the
⁢
⁢
k
th
⁢
⁢
class
⁢
⁢
of
customers
⁢
⁢
x
ij
=
{
1
,
for
⁢
⁢
the
⁢
⁢
i
th
⁢
⁢
commodity
⁢
⁢
being
⁢
⁢
placed
⁢
⁢
on
⁢
⁢
the
⁢
⁢
j
th
⁢
⁢
shelf
0
,
for
⁢
⁢
the
⁢
⁢
i
th
⁢
⁢
commodity
⁢
⁢
not
⁢
⁢
being
⁢
⁢
placed
⁢
⁢
on
⁢
⁢
the
⁢
⁢
j
th
⁢
⁢
shelf
.

### Claim 11 (depends on 7)

The system according to claim 7, further comprising RFID readers deployed at different locations in the supermarket, for detecting an RFID tag on a shopping cart used by a customer of the customers, wherein the shopping path obtaining device records dynamic locations of the shopping cart through the RFID reader, thereby deriving the shopping path of the customer.

### Claim 12 (depends on 7)

The system according to claim 7, wherein the classifying device classifies a customer of the customers based on one or more of following data sources:
a shopping list recorded by a point of sale (POS) machine; and
demographic of the customer.