Clickable Terms for Contextual Advertising

• US 20110246285A1
• Filed: 03/31/2010
• Published: 10/06/2011
• Est. Priority Date: 03/31/2010
• Status: Active Grant

• Current Assignee: R2 Solutions LLC (Acacia Research Corporation)
• Original Assignee: Yahoo! Inc. (Apollo Global Management, Inc.)
• Inventors: Ratnaparkhi, Adwait, Bagherjeiran, Abraham, Hatch, Andrew O.
• Granted Patent: US 8,533,043 B2
• Time in Patent Office: Days
• US Class Current: 705/14.42
• CPC Class Codes: G06Q 30/00 Commerce
• CPC Class Codes: G06Q 30/0243 Comparative campaigns

Abstract

An online advertising selects online advertisements for display on a network location taking into account a probability that a candidate online advertisement will receive a click on a particular website. The system may determine a network location identity of the network location and transform a set of advertisements into a set of ranked advertisements. The system may determine an advertisement rank of a first advertisement among the set of ranked advertisements. The system then may generate a click probability value. The click probability value may reflect a click probability of the first advertisement by dividing an exponent of a weighted sum of the network location identity and the advertisement rank by one plus the exponent of the weighted sum of the network location identity and the advertisement rank.

Claims

1. An online advertising method implemented in a computer to select advertisements for display on a network location, the method comprising:

determining, in a computer, a network location identity of the network location;

transforming a set of advertisements into a set of ranked advertisements;

determining, in the computer, an advertisement rank of a first advertisement among the set of ranked advertisements;

generating, in the computer, a click probability value that reflects a click probability of the first advertisement by dividing an exponent of a weighted sum of the network location identity and the advertisement rank by one plus the exponent of the weighted sum of the network location identity and the advertisement rank.

2. The method of claim 1, further comprising:

computing the click probability value according to the equation;

3. The method of claim 1, wherein generating the click probability value includes dividing an exponent of a weighted sum of the network location identity, the advertisement rank, and a set of ad-side terms in the first advertisement by one plus the exponent of the weighted sum of the network location identity, the advertisement rank, and the set of ad-side terms in the first advertisement.

4. The method of claim 3, wherein the click probability value is in a form Pr(c(a)|p, r(a), T(a)) to reflect the click probability value of a click event c(a) on the first advertisement a for the network location identity p, the advertisement rank r(a), and the set of ad-side terms T(a) in the first advertisement a, the method further comprising:

computing the weighted sum of the network location identity, the advertisement rank, and the set of ad-side terms in the first advertisement according to the equation;

5. The method of claim 3, further comprising:

computing the weighted sum of the network location identity, the advertisement rank, and the set of ad-side terms in the first advertisement according to the equation;

Σ

_j^wj*^fj*(p,r(a),T(a))=Σ

_iw_if_i(p,r(a))+Σ

_k<sup>w′

p,kf′

p,k</sup>, with

w′

_p,k˜

(_p,k,σ

)∀

p,k and

f′

_p,k≡

α

_p,k(T_kε

a)∀

p,k,

wherein Σ

_j^wj*^fj*(p, r(a), T(a)) is the weighted sum Σ

_j^wj* of the network location identity p, the advertisement rank r(a), and the set of ad-side terms T(a) in the first advertisement a, Σ

_iw_if_i(p, r(a)) is a weighted sum Σ

_i^wj* of the features f_i that include the network location identity p and the advertisement rank r(a), and Σ

_k<sup>w′

p,kf′

p,k</sup> is a weighted sum Σ

_k<sup>w′

p,k</sup> of an associated feature f′

_p,k that is defined as a term importance weight value α

_p,k for the ad-side term T_k on website identity p.

6. The method of claim 3, further comprising:

computing the weighted sum of the network location identity, the advertisement rank, and the set of ad-side terms in the first advertisement according to the equation;

7. The method of claim 3, further comprising:

computing the weighted sum of the network location identity, the advertisement rank, and the set of ad-side terms in the first advertisement according to the equation;

8. The method of claim 3, where the weighted sum of the network location identity, the advertisement rank, and the set of ad-side terms includes a term importance weight value, and where the term importance weight value is one of (i) equal to one, (ii) a function of a total number of unique ad side terms, and (iii) a function of an expected number of unique ad-side terms that co-occur with a given ad-side term in advertisements that appear on the network location.

9. The method of claim 8, further comprising:

computing the term importance weight value according to one of the following equations;

10. A computer readable medium containing executable instructions stored thereon, which, when executed in a computer, cause the computer to select online advertisements for display on a network location, the instructions for:

determining, in a computer, a network location identity of the network location;

transforming a set of advertisements into a set of ranked advertisements;

determining, in the computer, an advertisement rank of a first advertisement among the set of ranked advertisements;

generating, in the computer, a click probability value that reflects a click probability of the first advertisement by dividing an exponent of a weighted sum of the network location identity and the advertisement rank by one plus the exponent of the weighted sum of the network location identity and the advertisement rank.

11. The computer readable medium of claim 12, further comprising:

computing the click probability value according to the equation;

12. The computer readable medium of claim 10, wherein generating the click probability value includes dividing an exponent of a weighted sum of the network location identity, the advertisement rank, and a set of ad-side terms in the first advertisement by one plus the exponent of the weighted sum of the network location identity, the advertisement rank, and the set of ad-side terms in the first advertisement.

13. The computer readable medium of claim 12, wherein the click probability value is in a form Pr(c(a)|p, r(a), T(a)) to reflect the click probability value of a click event c(a) on the first advertisement a for the network location identity p, the advertisement rank r(a), and the set of ad-side terms T(a) in the first advertisement a, the method further comprising:

computing the weighted sum of the network location identity, the advertisement rank, and the set of ad-side terms in the first advertisement according to the equation;

14. The computer readable medium of claim 12, further comprising:

computing the weighted sum of the network location identity, the advertisement rank, and the set of ad-side terms in the first advertisement according to the equation;

Σ

_j^wj*^fj*(p,r(a),T(a))=Σ

_iw_if_i(p,r(a))+Σ

_k<sup>w′

p,kf′

p,k</sup>, with

w′

_p,k˜

(_p,k,σ

)∀

p,k and

f′

_p,k≡

α

_p,k(T_kε

a)∀

p,k,

wherein Σ

_j^wj*^fj*(p, r(a), T(a)) is the weighted sum Σ

_j^wj* of the network location identity p, the advertisement rank r(a), and the set of ad-side terms T(a) in the first advertisement a, Σ

_iw_if_i(p, r(a)) is a weighted sum Σ

_i^wi of the features f_i that include the network location identity p and the advertisement rank r(a), and Σ

_k<sup>w′

p,kf′

p,k</sup> is a weighted sum Σ

_k<sup>w′

p,k</sup> of an associated feature f′

_p,k that is defined as a term importance weight value α

_p,k for the ad-side term T_k on website identity p.

15. The computer readable medium of claim 12, further comprising:

computing the weighted sum of the network location identity, the advertisement rank, and the set of ad-side terms in the first advertisement according to the equation;

16. The computer readable medium of claim 12, further comprising:

computing the weighted sum of the network location identity, the advertisement rank, and the set of ad-side terms in the first advertisement according to the equation;

17. The computer readable medium of claim 12, wherein the weighted sum of the network location identity, the advertisement rank, and the set of ad-side terms comprises a term importance weight value, and wherein the term importance weight value comprises one of (i) equal to one, (ii) a function of a total number of unique ad side terms, and (iii) a function of an expected number of unique ad-side terms that co-occur with a given ad-side term in advertisements that appear on the network location.

18. The method of claim 17, further comprising:

computing the term importance weight value according to one of the following equations;

19. A system to select online advertisements for display on a network location, the system comprising:

at least one web server, comprising at least one processor and memory, to receive a set of advertisements; and

a processing and matching platform, comprising at least one processor and memory, coupled to the web server to determine a network location identity of the network location, transform the set of advertisements into a set of ranked advertisements, determine an advertisement rank of a first advertisement among the set of ranked advertisements, and generate a click probability value that reflects a click probability of the first advertisement by dividing an exponent of a weighted sum of the network location identity and the advertisement rank by one plus the exponent of the weighted sum of the network location identity and the advertisement rank.

20. The system of claim 19, the processing and matching platform further for computing the click probability value according to the equation:

21. The method of claim 19, where generating the click probability value includes dividing an exponent of a weighted sum of the network location identity, the advertisement rank, and a set of ad-side terms in the first advertisement by one plus the exponent of the weighted sum of the network location identity, the advertisement rank, and the set of ad-side terms in the first advertisement.

22. The system of claim 21, where the click probability value is in a form Pr(c(a)|p, r(a), T(a)) to reflect the click probability value of a click event c(a) on the first advertisement a for the network location identity p, the advertisement rank r(a), and the set of ad-side terms T(a) in the first advertisement a, the processing and matching platform further for computing the weighted sum of the network location identity, the advertisement rank, and the set of ad-side terms in the first advertisement according to the equation:

23. The system of claim 21, he processing and matching platform further for computing the weighted sum of the network location identity, the advertisement rank, and the set of ad-side terms in the first advertisement according to the equation:

Σ

_j^wj*^fj*(p,r(a),T(a))=Σ

_iw_if_i(p,r(a))+Σ

_k<sup>w′

p,kf′

p,k</sup>, with

w′

_p,k˜

(_p,k,σ

)∀

p,k and

f′

_p,k≡

α

_p,k(T_kε

a)∀

p,k,

wherein Σ

_j^wj*^fj*(p, r(a), T(a)) is the weighted sum Σ

_j^wj* of the network location identity p, the advertisement rank r(a), and the set of ad-side terms T(a) in the first advertisement a, Σ

_iw_if_i(p, r(a)) is a weighted sum Σ

_i^wi of the features f_i that include the network location identity p and the advertisement rank r(a), and Σ

_k<sup>w′

p,kf′

p,k</sup> is a weighted sum Σ

_k<sup>w′

p,k</sup> of an associated feature f′

_p,k that is defined as a term importance weight value α

_p,k for the ad-side term T_k on website identity p.

24. The system of claim 21, he processing and matching platform further for computing the weighted sum of the network location identity, the advertisement rank, and the set of ad-side terms in the first advertisement according to the equation:

25. The system of claim 21, he processing and matching platform further for computing the weighted sum of the network location identity, the advertisement rank, and the set of ad-side terms in the first advertisement according to the equation:

28. The system of claim 21, where the weighted sum of the network location identity, the advertisement rank, and the set of ad-side terms includes a term importance weight value, and where the term importance weight value is one of (i) equal to one, (ii) a function of a total number of unique ad side terms, and (iii) a function of an expected number of unique ad-side terms that co-occur with a given ad-side term in advertisements that appear on the network location.

27. The system of claim 26, he processing and matching platform further for computing the term importance weight value according to one of the following equations: