Keyword bidding strategy for novel concepts

• US 7,941,436 B2
• Filed: 11/30/2006
• Issued: 05/10/2011
• Est. Priority Date: 11/30/2006
• Status: Active Grant

• Current Assignee: R2 Solutions LLC (Acacia Research Corporation)
• Original Assignee: Yahoo! Inc. (Apollo Global Management, Inc.)
• Inventors: Dom, Byron, Brunk, Clifford A., Popescul, Alexandrin
• Primary Examiner(s): Stace; Brent
• Assistant Examiner(s): Rahman; Sabana
• Application Number:
  US11/607,299
• Publication Number:
  US 20080133503A1
• Time in Patent Office: 1,622 Days
• Field of Search: 707 1- 4
• US Class Current: 707/750
• CPC Class Codes: G06F 16/951 Indexing; Web crawling techniques
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising
• CPC Class Codes: G06Q 30/0275 Auctions
• CPC Class Codes: G06Q 30/0609 Buyer or seller confidence or verification
• CPC Class Codes: G06Q 30/08 Auctions

Abstract

A system and method are disclosed for identifying a keyword that is a novel concept or anomaly based on prior search results for the keyword. Advertisements may be sold for the keyword, or the keyword may be purchased or recommended for purchase based on anticipation of increased future searches on the keyword.

Claims

1. A method for utilizing an anomalous query in bidding comprising:

identifying a query from a plurality of search queries in a search log database;

monitoring the search log database to determine a first frequency at which the identified query has been searched for during a first time interval;

monitoring a second frequency at which the identified query is searched for during a second time interval;

comparing the first frequency with a second frequency, wherein the comparison comprises determining whether the second frequency increased over the first frequency above a certain amount that comprises a spike recall constant multiplied by a standard deviation, wherein the standard deviation is based on at least one earlier time interval and the spike recall constant represents a relative amount of frequency change; and

performing at least one of selling the identified query at an adjusted rate, purchasing the identified query for an advertisement, or recommending the identified query for an advertisement purchase when either the first frequency or the second frequency is above a threshold frequency and when the comparison indicates an increase above the certain amount.

2. The method according to claim 1 wherein the plurality of search queries are entered into a search engine and recorded to the search log database.

3. The method according to claim 1 wherein the at least one previous frequency comprises the first frequency and the first frequency comprises an average of a plurality of frequencies.

4. The method according to claim 1 wherein the certain amount is a percentage increase from the first frequency to the second frequency.

5. The method according to claim 1 wherein the monitoring of the search log database is in real-time.

6. The method according to claim 1 wherein the first frequency is a mean value of frequencies over the first time interval.

7. The method according to claim 1 wherein the comparison indicates an increase above the certain amount when the following equation is satisfied:

f/V−

μ

>

k*G, wherein f is the frequency of the identified keyword, V is the total number of keywords over a time period, μ

is mean of the normalized keyword frequency over the time period, k is the spike recall constant, and σ

is the standard deviation of the frequency normalized over the time period.

8. The method according to claim 1 wherein the second frequency is divided by a volume of queries when comparing with the first frequency.

9. The method according to claim 1 wherein the adjusted rate is an increased price based on the potential increase in frequency of the identified query.

10. In a computer readable storage medium having stored therein data representing instructions executable by a programmed processor for keyword identification, the storage medium comprising instructions for:

monitoring a keyword database that includes keyword usage data; and

identifying a keyword that has an unusually high frequency;

wherein the step of identifying comprises;

measuring a frequency of the identified keyword;

observing a frequency change with the keyword;

determining a threshold frequency; and

recognizing the keyword as having an unusually high frequency when the frequency exceeds the threshold frequency and when the frequency change exceeds a predetermined frequency change that is based on a spike recall constant multiplied by a standard deviation of at least one previous frequency, wherein the spike recall constant represents a relative amount of frequency change.

11. The storage medium according to claim 10 wherein the keyword usage data includes a record of search requests made for the keywords in the keyword database.

12. The storage medium according to claim 10 wherein the predetermined frequency change is a percentage increase in frequency over a predetermined time.

13. The storage medium according to claim 10 further comprising:

performing at least one of selling the identified keyword at an adjusted rate, purchasing the identified keyword for an advertisement, or recommending the identified keyword for an advertisement purchase.

14. The storage medium according to claim 13 wherein the performing step is only performed if the identified keyword is determined to have an unusually high frequency.

15. The storage medium according to claim 10 wherein the frequency change exceeds a predetermined frequency change if the following equation is satisfied:

f/V−

μ

>

k*σ

, wherein f is the frequency of the identified keyword, V is the total number of keywords over a time period, μ

is mean of the normalized keyword frequency over the time period, k is the spike recall constant, and σ

is the standard deviation of the frequency normalized over the time period.

16. The storage medium according to claim 15 wherein the spike recall constant represents a required increase in frequency for recognizing the keyword as having an unusually high frequency.

17. A system identifying an anomalous query comprising:

a search engine with a processor configured to receive queries and provide search results based on the queries;

a search log database coupled with the search engine and configured to record the queries received by the search engine; and

an event detector coupled with the search log database and configured to monitor the frequency of the queries that are input into the search engine, wherein the event detector is configured to recognize an anomalous query based on an increase in frequency of the anomalous query when the frequency exceeds a frequency threshold, and when a subsequent frequency divided by a volume of queries minus a prior frequency value is greater than a spike recall constant multiplied by a standard deviation.

18. The system according to claim 17 further comprising an ad server coupled with the search engine and the event detector.

19. The system according to claim 18 wherein the ad server provides at least one advertisement that is displayed by the search engine in response to receiving a query and providing the search result for that query.

20. The system according to claim 17 wherein if an anomalous query is recognized, the event detector is further configured to at least one of sell the anomalous query at an adjusted rate, purchase the anomalous query for an advertisement, or recommend the anomalous query for an advertisement purchase.