Keyword analysis using social media data

• US 10,540,660 B1
• Filed: 05/28/2010
• Issued: 01/21/2020
• Est. Priority Date: 05/19/2010
• Status: Active Grant

• Current Assignee: Adobe Inc.
• Original Assignee: Adobe Inc.
• Inventors: Smith, Kevin G., George, William Brandon
• Primary Examiner(s): Bekerman, Michael
• Assistant Examiner(s): Snider, Scott
• Application Number:
  US12/789,735
• Time in Patent Office: 3,525 Days
• Field of Search: 705 141
• CPC Class Codes: G06Q 30/00 Commerce
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising
• CPC Class Codes: G06Q 30/0256 User search
• CPC Class Codes: G06Q 50/01 Social networking
• CPC Class Codes: H04L 51/52 for supporting social networking services

Abstract

A system and method is disclosed for bidding on search engine marketing keywords at a search engine. The system comprises a bid management system, which is configured to bid on keywords at a search engine, configured to determine a keyword for evaluation and one or more measures for the keyword. The measures are dependent on social media content hosted by one or more social media platforms. The bid management system uses the social media measures to determine a level of popularity for the keyword based at least in part on the one or more measures.

Claims

1. A method that includes a bid management server system having one or more processing devices performing operations comprising:

identifying a target keyword for evaluation;

controlling a display of search results by a search engine when the search engine services a query including the target keyword, wherein controlling the display of the search results comprises;

establishing communication sessions with a plurality of social media platforms via respective web application programming interfaces of the social media platforms,

retrieving, from the plurality of social media platforms and during the communication sessions, social media metrics for the target keyword,

determining that a spike has occurred for the target keyword, wherein the spike comprises an increase in the social media metrics of the target keyword that exceeds a threshold increase included in a policy of the bid management server system;

computing, responsive to determining that the spike has occurred, a bid for the target keyword, wherein computing the bid comprises;

(i) selecting comparison keywords having comparison social media metrics similar to the retrieved social media metrics, wherein the comparison social media metrics are dependent on measurements of post activity associated with the plurality of social media platforms,

(ii) correlating, based on a statistical analysis, the comparison social media metrics with comparison search engine marketing metrics of the comparison keywords, and

(iii) generating the bid from the comparison search engine marketing metrics, and

transmitting the bid to a web server hosting the search engine, wherein transmitting the bid causes the search engine to position, in the search results and based on the transmitted bid, content associated with the bid management server system.

2. The method of claim 1, wherein identifying the target keyword for evaluation comprises:

identifying a particular electronic post on one or more social media platforms;

extracting one or more search terms from metadata of the particular electronic post;

matching the one or more search terms to metadata in additional electronic posts on the one or more social media platforms;

creating a filtered set of additional electronic posts from the additional electronic posts by excluding a subset of the additional electronic posts based on metadata in the subset of the additional electronic posts indicating a lack of relevance to the particular electronic post, the lack of relevance comprising one or more of (i) the metadata in the subset of the additional electronic posts identifying a posting time subsequent to the particular electronic post or (ii) the metadata in the subset of the additional electronic posts lacking a network link to the particular electronic post, wherein the filtered set of additional electronic posts includes the target keyword; and

selecting the target keyword based on social media metrics for the target keyword that are computed from the filtered set of additional electronic posts.

3. The method of claim 2, wherein selecting the target keyword based on the social media metrics for the target keyword comprises:

identifying a set of candidate keywords by (i) parsing the filtered set of additional electronic posts and (ii) excluding, from results of the parsing, a set of defined exclusion words;

computing a set of relevance metric values for the set of candidate keywords, respectively; and

selecting the target keyword from the set of candidate keywords based on a computed relevance metric value for the target keyword from the set of relevance metric values.

4. The method of claim 3, wherein computing each relevance metric value comprises ranking each candidate keyword based according to one or more of:

a number of times the candidate keyword occurs in the filtered set of additional electronic posts;

a number of unique posts or threads that includes the candidate keyword,

a size of an audience that has viewed at least one post from the filtered set of additional electronic posts containing the candidate keyword;

or

a contributor weight assigned to an author associated with a contributor computing device that has posted electronic content in the filtered set of additional electronic posts having the candidate keyword.

5. The method of claim 2, wherein the social media metrics for the target keyword and the retrieved social media metrics comprise one or more of:

a number of direct replies or downstream replies published to the plurality of social media platforms;

a number of unique contributors to posts published to the plurality of social media platforms;

a number of unique viewers of the posts published to the plurality of social media platforms;

sentiment information computed for the posts published to the plurality of social media platforms; and

forwarding velocity or reply velocity for the posts published to the plurality of social media platforms.

6. The method of claim 1, wherein identifying the target keyword for evaluation comprises:

identifying a particular electronic post on one or more social media platforms;

determining identifying information associated with an author of the particular electronic post;

identifying additional electronic posts from the author on the one or more social media platforms by searching the one or more social media platforms for content that is posted by the author with an account associated with at least a portion of the identifying information;

identifying downstream content associated with the content posted by the author;

determining a first reposting rate at which the content posted by the author is subsequently also posted by other authors within a reposting time period on the one or more social media platforms, the downstream content comprising the content subsequently also posted by the authors;

comparing the first reposting rate of the author to a second reposting rate of one of the other authors within the reposting time period, the second reposting rate determined based on the downstream content;

determining that the first reposting rate of the author is higher than the second reposting rate of the one of the other authors;

assigning, based on the first reposting rate being higher than the second reposting rate, a contributor weight to the author for the downstream content that is higher relative to an additional contributor weight assigned to the one of the other authors for the downstream content, wherein the additional contributor weight is assigned to the one of the other authors based on the second reposting rate; and

selecting, based on the author having the contributor weight, the target keyword from one or more of (i) particular electronic post and (ii) the additional electronic posts.

7. The method of claim 6, wherein determining the identifying information associated with the author comprises:

obtaining first identification information elements through first communications via a first application programming interface for a first social media platform at which the particular electronic post is posted, wherein the first identification information elements comprise one or more of a legal name of the author, a physical address of the author, and a photograph of the author;

querying, through second communication with a second social media platform via a second application programming interface for, the second social media platform with a combination of a subset of the first identification information elements; and

receiving, through the second communications and responsive to the querying, second identification information elements;

aggregating, into the identifying information, the first identification information elements and the second identification information elements, wherein the aggregating is based on a statistical analysis of the first identification information elements and the second identification information elements indicating a threshold probability of the second identification information elements belonging to the author.

8. A non-transitory computer-readable storage medium having instructions stored thereon that, when executed by a processor of a bid management server system, cause the processor to perform operations comprising:

identifying a target keyword for evaluation;

controlling a display of search results by a search engine when the search engine services a query including the target keyword, wherein controlling the display of the search results comprises;

establishing communication sessions with a plurality of social media platforms via respective web application programming interfaces of the social media platforms,

retrieving, from the plurality of social media platforms and during the communication sessions, social media metrics for the target keyword,

determining that a spike has occurred for the target keyword, wherein the spike comprises an increase in the social media metrics of the target keyword that exceeds a threshold increase included in a policy of the bid management server sy stem;

computing, responsive to determining that the spike has occurred, a bid for the target keyword, wherein computing the bid comprises;

(i) selecting comparison keywords having comparison social media metrics similar to the retrieved social media metrics, wherein the comparison social media metrics are dependent on measurements of post activity associated with the plurality of social media platforms,

(ii) correlating, based on a statistical analysis, the comparison social media metrics with comparison search engine marketing metrics of the comparison keywords, and

(iii) generating the bid from the comparison search engine marketing metrics, and

transmitting the bid to a web server hosting the search engine, wherein transmitting the bid causes the search engine to position, in the search results and based on the transmitted bid, content associated with the bid management server system.

9. The non-transitory computer-readable storage medium of claim 8, wherein identifying the target keyword for evaluation comprises:

identifying a particular electronic post on one or more social media platforms;

extracting one or more search terms from metadata of the particular electronic post;

matching the one or more search terms to metadata in additional electronic posts on the one or more social media platforms;

creating a filtered set of additional electronic posts from the additional electronic posts by excluding a subset of the additional electronic posts based on metadata in the subset of the additional electronic posts indicating a lack of relevance to the particular electronic post, the lack of relevance comprising one or more of (i) the metadata in the subset of the additional electronic posts identifying a posting time subsequent to the particular electronic post or (ii) the metadata in the subset of the additional electronic posts lacking a network link to the particular electronic post, wherein the filtered set of additional electronic posts includes the target keyword; and

selecting the target keyword based on social media metrics for the target keyword that are computed from the filtered set of additional electronic posts.

10. The non-transitory computer-readable storage medium of claim 9, wherein selecting the target keyword based on the social media metrics for the target keyword comprises:

identifying a set of candidate keywords by (i) parsing the filtered set of additional electronic posts and (ii) excluding, from results of the parsing, a set of defined exclusion words;

computing a set of relevance metric values for the set of candidate keywords, respectively; and

selecting the target keyword from the set of candidate keywords based on a computed relevance metric value for the target keyword from the set of relevance metric values.

11. The non-transitory computer-readable storage medium of claim 10, wherein computing each relevance metric value comprises ranking each candidate keyword based according to one or more of:

a number of times the candidate keyword occurs in the filtered set of additional electronic posts;

a number of unique posts or threads that includes the candidate keyword,

a size of an audience that has viewed at least one post from the filtered set of additional electronic posts containing the candidate keyword;

or

a contributor weight assigned to an author associated with a contributor computing device that has posted electronic content in the filtered set of additional electronic posts having the candidate keyword.

12. The non-transitory computer-readable storage medium of claim 11, wherein computing the relevance metric value comprises computing the contributor weight assigned to an author of the particular electronic post wherein computing the contributor weight comprises:

determining identifying information associated with the author;

identifying additional electronic posts from the author on the one or more social media platforms by searching the one or more social media platforms for content that is posted by the author with an account associated with at least a portion of the identifying information;

identifying downstream content associated with the content posted by the author;

determining a first reposting rate at which the content posted by the author is subsequently also posted by other authors within a reposting time period on the one or more social media platforms, the downstream content comprising the content subsequently also posted by the authors;

comparing the first reposting rate of the author to a second reposting rate of one of the other authors within the reposting time period, the second reposting rate determined based on the downstream content;

determining that the first reposting rate of the author is higher than the second reposting rate of the one of the other authors; and

assigning, based on the first reposting rate being higher than the second reposting rate, the contributor weight to the author for the downstream content that is higher relative to an additional contributor weight assigned to the one of the other authors for the downstream content, wherein the additional contributor weight is assigned to the one of the other authors based on the second reposting rate.

13. The non-transitory computer-readable storage medium of claim 12, wherein determining the identifying information associated with the author comprises:

obtaining first identification information elements through first communications via a first application programming interface for a first social media platform at which the particular electronic post is posted, wherein the first identification information elements comprise one or more of a legal name of the author, a physical address of the author, and a photograph of the author;

querying, through second communication with a second social media platform via a second application programming interface for, the second social media platform with a combination of a subset of the first identification information elements; and

receiving, through the second communications and responsive to the querying, second identification information elements;

aggregating, into the identifying information, the first identification information elements and the second identification information elements, wherein the aggregating is based on a statistical analysis of the first identification information elements and the second identification information elements indicating a threshold probability of the second identification information elements belonging to the author.

14. A system comprising:

a search engine server configured for;

receiving a query from a client device via a data network,

servicing the query by retrieving web content results matching a query parameter having a target keyword,

positioning, based on bid data, one of the web content results in an ordered list of the web content results,

generating a graphical interface that displays the ordered list of the web content results, and

transmitting the graphical interface to the client device;

a plurality of social media servers configured for;

creating, responsive to communication with subscriber client devices, electronic posts and electronic responsive posts, wherein the electronic posts and the electronic responsive posts have metadata describing reply-based relationships among the electronic posts and the electronic responsive posts, and

a bid management server system configured for providing an interface between the search engine server and the plurality of social media servers, wherein the bid management server system is communicatively coupled to the search engine server and the social media servers, the bid management server system comprising;

a memory having instructions stored thereon;

a processing device configured for executing the instructions and thereby performing operations comprising;

identifying the target keyword for evaluation,

establishing communication sessions with the plurality of social media servers via respective web application programming interfaces of a plurality of social media platforms hosted by the plurality of social media servers,

retrieving, from the plurality of social media platforms and during the communication sessions, social media metrics for the target keyword based on at least some of the electronic posts and the electronic responsive posts,

determining that a spike has occurred for the target keyword, wherein the spike comprises an increase in the social media metrics of the target keyword that exceeds a threshold increase included in a policy of the bid management server system,

computing, responsive to determining that the spike has occurred, a bid for the target keyword, wherein computing the bid comprises;

(i) selecting comparison keywords having comparison social media metrics similar to the retrieved social media metrics, wherein the comparison social media metrics are dependent on measurements of post activity associated with the plurality of social media platforms,

(ii) correlating, based on a statistical analysis, the comparison social media metrics with comparison search engine marketing metrics of the comparison keywords, and

(iii) generating the bid from the comparison search engine marketing metrics, and

a network interface device configured for transmitting the bid data to the search engine server.

15. The system of claim 14, wherein identifying the target keyword for evaluation comprises:

identifying a particular electronic post on one or more social media platforms;

extracting one or more search terms from metadata of the particular electronic post;

matching the one or more search terms to metadata in additional electronic posts on the one or more social media platforms;

creating a filtered set of additional electronic posts from the additional electronic posts by excluding a subset of the additional electronic posts based on metadata in the subset of the additional electronic posts indicating a lack of relevance to the particular electronic post, the lack of relevance comprising one or more of (i) the metadata in the subset of the additional electronic posts identifying a posting time subsequent to the particular electronic post or (ii) the metadata in the subset of the additional electronic posts lacking a network link to the particular electronic post, wherein the filtered set of additional electronic posts includes the target keyword; and

selecting the target keyword based on social media metrics for the target keyword that are computed from the filtered set of additional electronic posts.

16. The system of claim 15, wherein selecting the target keyword based on the social media metrics for the target keyword comprises:

identifying a set of candidate keywords by (i) parsing the filtered set of additional electronic posts and (ii) excluding, from results of the parsing, a set of defined exclusion words;

computing a set of relevance metric values for the set of candidate keywords, respectively; and

selecting the target keyword from the set of candidate keywords based on a computed relevance metric value for the target keyword from the set of relevance metric values.

17. The system of claim 16, wherein computing each relevance metric value comprises ranking each candidate keyword based according to one or more of:

a number of times the candidate keyword occurs in the filtered set of additional electronic posts;

a number of unique posts or threads that includes the candidate keyword,

a size of an audience that has viewed at least one post from the filtered set of additional electronic posts containing the candidate keyword;

or

a contributor weight assigned to a contributor computing device that has posted electronic content in the filtered set of additional electronic posts having the candidate keyword.

18. The system of claim 15, wherein the social media metrics for the target keyword and the retrieved social media metrics comprise one or more of:

a number of direct replies or downstream replies published to the plurality of social media platforms;

a number of unique contributors to posts published to the plurality of social media platforms;

a number of unique viewers of the posts published to the plurality of social media platforms;

sentiment information computed for the posts published to the plurality of social media platforms; and

forwarding velocity or reply velocity for the posts published to the plurality of social media platforms.

19. The system of claim 14, wherein identifying the target keyword for evaluation comprises:

identifying a particular electronic post on one or more social media platforms;

determining identifying information associated with an author of the particular electronic post;

identifying additional electronic posts from the author on the one or more social media platforms by searching the one or more social media platforms for content that is posted by the author with an account associated with at least a portion of the identifying information;

identifying downstream content associated with the content posted by the author;

determining a first reposting rate at which the content posted by the author is subsequently also posted by other authors within a reposting time period on the one or more social media platforms, the downstream content comprising the content subsequently also posted by the authors;

comparing the first reposting rate of the author to a second reposting rate of one of the other authors within the reposting time period, the second reposting rate determined based on the downstream content;

determining that the first reposting rate of the author is higher than the second reposting rate of the one of the other authors;

assigning, based on the first reposting rate being higher than the second reposting rate, a contributor weight to the author for the downstream content that is higher relative to an additional contributor weight assigned to the one of the other authors for the downstream content, wherein the additional contributor weight is assigned to the one of the other authors based on the second reposting rate; and

selecting, based on the author having the contributor weight, the target keyword from one or more of (i) particular electronic post and (ii) the additional electronic posts.

20. The system of claim 19, wherein determining the identifying information associated with the author comprises:

obtaining first identification information elements through first communications via a first application programming interface for a first social media platform at which the particular electronic post is posted, wherein the first identification information elements comprise one or more of a legal name of the author, a physical address of the author, and a photograph of the author;

querying, through second communication with a second social media platform via a second application programming interface for, the second social media platform with a combination of a subset of the first identification information elements; and

receiving, through the second communications and responsive to the querying, second identification information elements;

aggregating, into the identifying information, the first identification information elements and the second identification information elements, wherein the aggregating is based on a statistical analysis of the first identification information elements and the second identification information elements indicating a threshold probability of the second identification information elements belonging to the author.