Behavioral targeting system that generates user profiles for target objectives
A behavioral targeting system determines user profiles from online activity. The system includes a plurality of models that define parameters for determining a user profile score. Event information, which comprises on-line activity of the user, is received at an entity. To generate a user profile score, is model is selected. The model comprises recency, intensity and frequency dimension parameters. The behavioral targeting system generates a user profile score for a target objective, such as brand advertising or direct response advertising. The parameters from the model are applied to generate the user profile score in a category. The behavioral targeting system has application for use in ad serving to on-line users.
1. A method comprising:
receiving, by a processor, a search query from a user;
parsing, by the processor, the search query to obtain one or more query terms;
retrieving, by the processor from a database, a plurality of tokens matching the query terms;
identifying, by the processor, a plurality of categories, each of the plurality of identified categories being associated with a retrieved token;
assigning, by the processor, an ambiguity parameter value to each retrieved token, the ambiguity parameter value being a factor of a conditional probability of an identified category associated with a corresponding token being an overall category for the search query given a presence of the corresponding token within the search query;
selecting, by the processor, a dominant token from the plurality of retrieved tokens based on one or more associated categories in the plurality of identified categories, based on a corresponding assigned ambiguity parameter value, and based on rules designed to rank the plurality of retrieved tokens; and
categorizing, by the processor, the search query based on the one or more categories associated with the dominant token.
2. The method of claim 1 , further comprising selecting, by the processor, the plurality of retrieved tokens having a highest probability to appear within the search query.
3. The method of claim 2 , wherein the selecting is based on a unit frequency parameter associated with each token, which specifies how many times each token is contained within the search query.
4. The method of claim 1 , further comprising selecting, by the processor, the plurality of retrieved tokens being longest possible tokens.
5. The method of claim 4 , wherein the longest possible tokens comprise tokens having a greatest number of words.
6. The method of claim 4 , wherein the longest possible tokens comprise tokens having a greatest length.
7. The method of claim 1 , wherein the selecting of the dominant token further comprises selecting the dominant token based on rules specifying elimination of tokens that contain stop words.
8. The method of claim 1 , wherein the selecting of the dominant token further comprises selecting the dominant token based on rules specifying a minimum frequency of token appearance within stored search queries.
9. A computing device comprising:
a hardware processor;
a non-transitory computer-readable storage medium for tangibly storing thereon program logic for execution by the hardware processor, the stored program logic comprising:
receiving logic, executed by the hardware processor, for receiving a search query from a user;
parsing logic, executed by the hardware processor, for parsing the search query to obtain one or more query terms;
retrieving logic, executed by the hardware processor, for retrieving, from a database, a plurality of tokens matching the query terms;
identifying logic, executed by the hardware processor, for identifying a plurality of categories, each of the plurality of identified categories being associated with a retrieved token;
assigning logic, executed by the hardware processor, for assigning an ambiguity parameter value to each retrieved token, the ambiguity parameter value being a factor of a conditional probability of an identified category associated with a corresponding token being an overall category for the search query given a presence of the corresponding token within the search query;
selecting logic, executed by the hardware processor, for selecting a dominant token from the plurality of retrieved tokens based on one or more associated categories in the plurality of identified categories, based on a corresponding assigned ambiguity parameter value, and based on rules designed to rank the plurality of retrieved tokens; and
categorizing logic, executed by the hardware processor, for categorizing the search query based on the one or more categories associated with the dominant token.
10. The computing device of claim 9 , further comprising second selecting logic executed by the processor for selecting the plurality of retrieved tokens having a highest probability to appear within the search query.
11. The computing device of claim 10 , wherein the second selecting logic comprising selecting based on a unit frequency parameter associated with each token, which specifies how many times each token is contained within the search query.
12. The computing device of claim 9 , further comprising second selecting logic executed by the processor for selecting the plurality of retrieved tokens being longest possible tokens.
13. The computing device of claim 12 , wherein the longest possible tokens comprise tokens having a greatest number of words.
14. The computing device of claim 12 , wherein the longest possible tokens comprise tokens having a greatest length.
15. The computing device of claim 9 , wherein the selecting logic for selecting the dominant token further comprises selecting the dominant token based on rules specifying elimination of tokens that contain stop words.
16. The computing device of claim 9 , wherein the selecting logic for selecting the dominant token further comprises selecting the dominant token based on rules specifying a minimum frequency of token appearance within stored search queries.
17. A non-transitory computer readable storage medium tangibly encoded with program code, that when executed by a processor of a computing device, causes the processor to execute instructions to:
receive, by a processor, a search query from a user;
parse, by the processor, the search query to obtain one or more query terms;
retrieve, by the processor from a database, a plurality of tokens matching the query terms;
identify, by the processor, a plurality of categories, each of the plurality of identified categories being associated with a retrieved token;
assign, by the processor, an ambiguity parameter value to each retrieved token, the ambiguity parameter value being a factor of a conditional probability of an identified category associated with a corresponding token being an overall category for the search query given a presence of the corresponding token within the search query;
select, by the processor, a dominant token from the plurality of retrieved tokens based on one or more associated categories in the plurality of identified categories, based on a corresponding assigned ambiguity parameter value, and based on rules designed to rank the plurality of retrieved tokens; and
categorize, by the processor, the search query based on the one or more categories associated with the dominant token.
18. The non-transitory computer readable storage medium of claim 17 , further comprising instructions comprising:
selecting, by the processor, the plurality of retrieved tokens having a highest probability to appear within the search query.