Systems and methods for control of event rates for segmented online campaigns

Patent metadata

Field	Value
Publication number	US-11636508-B2
Application number	US-201615369556-A
Country	US
Kind code	B2
Filing date	Dec 5, 2016
Priority date	Dec 5, 2016
Publication date	Apr 25, 2023
Grant date	Apr 25, 2023

Abstract

Allocating bids for providing content within a segmented campaign is controlled to ensure that an event rate associated with the provided content meets or exceeds a threshold rate. A campaign-level event rate, associated with the provided content, is estimated and provided as a feedback signal. This feedback signal is employed to dynamically update bid allocations for each of the segments, which in turn varies the number or rate of provided impressions and events. Such feedback enables the control of the campaign-level rate and ensures that the campaign-level rate meets or exceeds the rate threshold. To control these rates, the number of total impressions, as well as the number of associated events, is temporally sampled across the campaign segments. Based on the number of impressions and events, the campaign-level event rate is estimated and employed as the feedback signal. Updating the bid allocations may be based on the Beta Distribution.

First claim

What is claimed is:

1. A computer-implemented method comprising:
determine a control signal based on a comparison between a rate threshold and a campaign-level rate-estimate signal that corresponds to a probability that content previously provided in an online campaign is associated with an event;
segmenting the online campaign into a plurality of campaign segments comprising a first campaign segment and a second campaign segment;
determine one or more allocation signals comprising:
(i) a first allocation signal, for the first campaign segment, indicative of a first probability of participating in bidding to provide subsequent content in the online campaign for the first campaign segment; and
(ii) a second allocation signal, for the second campaign segment, indicative of a second probability of participating in bidding to provide the subsequent content in the online campaign for the second campaign segment, wherein the second probability is different than the first probability;

updating the one or more allocation signals to generate one or more updated bid allocation signals based on the control signal and a function that provides a continuous relationship between the control signal and each of the one or more allocation signals, wherein generating the one or more updated bid allocation signals comprises:
(i) updating the first allocation signal by employing a continuous function, corresponding to a relationship between of the control signal, one or more shape parameters, one or more configuration parameters, and a first segment-level rate-estimate signal for the first campaign segment, to generate a first updated bid allocation signal, for the first campaign segment, indicative of a first updated probability of participating in bidding to provide the subsequent content in the online campaign; and
(ii) updating the second allocation signal by employing the continuous function, corresponding to a relationship between the control signal, the one or more shape parameters, the one or more configuration parameters, and a second segment-level rate-estimate signal for the second campaign segment, to generate a second updated bid allocation signal, for the second campaign segment, indicative of a second updated probability of participating in bidding to provide subsequent content in the online campaign; and

providing the subsequent content in the online campaign based on the one or more updated bid allocation signals,
wherein at least one of the determining the one or more allocation signals or the updating the one or more allocation signals is based on a ratio of:
an incomplete beta function based upon (i) time, (ii) a first shape parameter and (iii) a second shape parameter; to
a second beta function of (i) the first shape parameter, for an actuator configuration parameter, of the control signal and (ii) the second shape parameter, for the actuator configuration parameter, of the control signal.

2. The computer-implemented method of claim 1 , further comprising:
receiving one or more input signals that are generated in response to providing the subsequent content in the online campaign, wherein the one or more input signals correspond to a number of events associated with the subsequent content and a number of impressions associated with the subsequent content;
updating the campaign-level rate-estimate signal to generate an updated campaign-level rate-estimate signal based on the one or more input signals;
updating the control signal to generate an updated control signal based on a comparison between the rate threshold and the updated campaign-level rate-estimate signal; and
updating the one or more allocation signals based on the updated control signal and the function.

3. The computer-implemented method of claim 1 , wherein the online campaign includes a plurality of campaign segments and updating the one or more allocation signals is further based on a plurality of segment-level rate-estimate signals, wherein each of the plurality of segment-level rate-estimate signals corresponds to one of the plurality of campaign segments and an event rate for the one of the plurality of campaign segments.

4. The computer-implemented method of claim 3 , wherein the function further provides a continuous relationship between the plurality of segment-level rate-estimate signals and each of one or more segment-level allocation signals.

5. The computer-implemented method of claim 1 , wherein the function is based on a continuous probability distribution.

6. The computer-implemented method of claim 1 , wherein the function includes an incomplete beta function.

7. The computer-implemented method of claim 1 , further comprising:
receiving a feedback signal based on the campaign-level rate-estimate signal; and
controlling the campaign-level rate-estimate signal based on the feedback signal such that the controlled campaign-level rate-estimate signal is greater than or equal to the rate threshold.

8. The computer-implemented method of claim 7 , wherein controlling the campaign-level rate-estimate signal includes employing a feedback based controller.

9. The computer-implemented method of claim 1 , wherein at least one of the determining the one or more allocation signals or the updating the one or more allocation signals is based on the equation:

a
i

(

p
ˆ

i

,
u

)

=

1

B
⁡
(

α
c

(
u
)

,

β
c

(
u
)

)

⁢

∫
0

p
i

ˆ

t

α
-
1

(

1
-
t

)

β
-
1

⁢
dt

Assignees

• Oath Americas Inc
• Yahoo Ad Tech Llc

Inventors

• Karlsson Niklas
• Sang qian
• Guo Jiaxing

Classifications

• G06Q30/0277

  Online advertisement · CPC title

• G06Q30/0244 Primary

  Optimization · CPC title

• G06Q30/0275

  Auctions · CPC title