Method and system for generating a linear machine learning model for predicting online user input actions
View Patent ↗A method of targeting receives several granular events and preprocesses the received granular events thereby generating preprocessed data to facilitate construction of a model based on the granular events. The method generates a predictive model by using the preprocessed data. The predictive model is for determining a likelihood of a user action. The method trains the predictive model. A system for targeting includes granular events, a preprocessor for receiving the granular events, a model generator, and a model. The preprocessor has one or more modules for at least one of pruning, aggregation, clustering, and/or filtering. The model generator is for constructing a model based on the granular events, and the model is for determining a likelihood of a user action. The system of some embodiments further includes several users, a selector for selecting a particular set of users from among the several users, a trained model, and a scoring module.
1. A computer implemented method of behavioral targeting comprising:
receiving, at a computer, a plurality of granular events, wherein a granular event comprises an on-line activity between a user and an entity;
preprocessing, using a computer, the received granular events to determine informational content of each granular event;
generating, in a computer, preprocessed data comprising a plurality of input features by grouping the granular events into a plurality of clusters based on the informational content, wherein each cluster corresponds to a unique input feature from among the input features; and
generating, in a computer, a predictive model based on a linear machine-learning model, the predictive model for determining, using the clusters and a linear combination of their corresponding input features, a likelihood of a predicted input action by a user.
2. The computer implemented method of claim 1 wherein the linear machine-learning model is at least one of a Support Vector Machine (SVM), a Naïve Bayes machine, Maximum Entropy, a logistic regression, and a linear regression model.
3. The computer implemented method of claim 1 , wherein the granular event comprises at least one of viewing a web page, clicking on a link in the web page, clicking on an advertisement in the web page, issuing a search query, using a search engine, filling out a form, posting, rating a page, rating a product, and performing a transaction.
4. The computer implemented method of claim 1 , wherein the online activity comprises a search and the method further comprises tracking a number of clicks on a search result.
5. The computer implemented method of claim 1 , wherein the online activity comprises a page view and the method further comprises counting, for each page of a plurality of pages, a number of page views.
6. The computer implemented method of claim 1 , wherein the informational content comprises at least one of search, search-click, sponsored search-click, page view, advertisement view, and advertisement click.
7. The computer implemented method of claim 1 , wherein the preprocessing comprises at least one of pruning, aggregating, and filtering the received granular events.
8. The computer implemented method of claim 1 , further comprising scoring the user based on a linear combination of granular event counts from a behavioral history.
9. The computer implemented method of claim 1 , further comprising scoring the user based on a ratio between an estimated probability of being a clicker and an estimated probability of being a non-clicker.
10. The computer implemented method of claim 1 , further comprising:
ranking the user in relation to other users based on a plurality of user scores;
generating a set of ranked users; and
identifying a subset of ranked users for additional targeting steps.
11. A non-transitory computer readable medium that stores a set of instructions which, when executed by a computer, cause the computer to execute steps for behavioral targeting, the steps comprising:
receiving a plurality of granular events, wherein a granular event comprises an on-line activity between a user and an entity;
preprocessing the received granular events to determine informational content of each granular event;
generating preprocessed data comprising a plurality of input features by grouping the granular events into a plurality of clusters based on the informational content, wherein each cluster corresponds to a unique input feature from among the input features; and
generating a predictive model based on a linear machine-learning model, the predictive model for determining, using the clusters and a linear combination of their corresponding input features, a likelihood of a predicted input action by a user.
12. The computer readable medium of claim 11 wherein the linear machine-learning model is at least one of a Support Vector Machine (SVM), a Naïve Bayes machine, Maximum Entropy, a logistic regression, and a linear regression model.
13. The computer readable medium of claim 11 , wherein the granular event comprises at least one of viewing a web page, clicking on a link in the web page, clicking on an advertisement in the web page, issuing a search query, using a search engine, filling out a form, posting, rating a page, rating a product, and performing a transaction.
14. The computer readable medium of claim 11 , wherein the online activity comprises a search and the method further comprises tracking a number of clicks on a search result.
15. The computer readable medium of claim 11 , wherein the online activity comprises a page view and the method further comprises counting, for each page of a plurality of pages, a number of page views.
16. The computer readable medium of claim 11 , wherein the informational content comprises at least one of search, search-click, sponsored search-click, page view, advertisement view, and advertisement click.
17. The computer readable medium of claim 11 , wherein the preprocessing comprises at least one of pruning, aggregating, and filtering the received granular events.
18. The computer readable medium of claim 11 , further comprising scoring the user based on a linear combination of granular event counts from a behavioral history.
19. The computer readable medium of claim 11 , further comprising scoring the user based on a ratio between an estimated probability of being a clicker and an estimated probability of being a non-clicker.
20. The computer readable medium of claim 11 , further comprising:
ranking the user in relation to other users based on a plurality of user scores;
generating a set of ranked users; and
identifying a subset of ranked users for additional targeting steps.