SYSTEMS AND METHODS FOR PAGE RECOMMENDATIONS BASED ON ONLINE USER BEHAVIOR
Systems, methods, and non-transitory computer readable media configured to determine features based on online user behavior regarding a seed content item and a candidate content item that may be presented in response to an indication of approval by a user regarding the seed content item. The features are processed to generate a probability that the user will interact with the candidate content item. The candidate content item is selected for presentation to the user based on the probability that the user will interact with the candidate content item.
1 . A computer-implemented method comprising:
determining, by a computing system, features based on online user behavior regarding a seed content item and a candidate content item that may be presented in response to an indication of approval by a user regarding the seed content item;
processing, by the computing system, the features to generate a probability that the user will interact with the candidate content item; and
selecting, by the computing system, the candidate content item for presentation to the user based on the probability that the user will interact with the candidate content item.
2 . The computer-implemented method of claim 1 , wherein the seed content item is a seed page and the candidate content item is a candidate page.
3 . The computer-implemented method of claim 2 , wherein the features include at least one of a number of conversions for each of the seed page, the candidate page, and a seed page/candidate page pair; a number of impressions for each of the seed page, the candidate page, and the seed page/candidate page pair; a ratio of a conversions rate for the seed page, a conversion rate for the candidate page, and a conversion rate for the seed page/candidate page pair; and a ratio of a conversion rate for the seed page/candidate page pair over a conversation rate for the seed page and a ratio of a conversion rate for the seed page/candidate page pair over a conversion rate for the candidate page.
4 . The computer-implemented method of claim 1 , wherein the selecting the candidate content item for presentation to the user based on the probability that the user will interact with the candidate content item comprises:
comparing the probability that the user will interact with the candidate content item with other probabilities associated with other candidate content items.
5 . The computer-implemented method of claim 1 , further comprising categorizing the features based on demographic categories.
6 . The computer-implemented method of claim 5 , wherein the demographic categories include at least one of gender, age category, and country.
7 . The computer-implemented method of claim 1 , further comprising applying a smoothing constant value to at least one feature of the features.
8 . The computer-implemented method of claim 1 , wherein the processing the features to generate a probability that a user will interact with the candidate content item comprises:
applying a machine learning technique based on the features.
9 . The computer-implemented method of claim 1 , wherein the machine learning technique includes at least one boosted decision tree algorithm.
10 . The computer-implemented method of claim 1 , wherein at least one of the seed content item and the candidate content item is associated with a page of a social networking system.
11 . A system comprising:
at least one processor; and
a memory storing instructions that, when executed by the at least one processor, cause the system to perform:
determining features based on online user behavior regarding a seed content item and a candidate content item that may be presented in response to an indication of approval by a user regarding the seed content item;
processing the features to generate a probability that the user will interact with the candidate content item; and
selecting the candidate content item for presentation to the user based on the probability that the user will interact with the candidate content item.
12 . The system of claim 11 , wherein the seed content item is a seed page and the candidate content item is a candidate page.
13 . The system of claim 12 , wherein the features include at least one of a number of conversions for each of the seed page, the candidate page, and a seed page/candidate page pair; a number of impressions for each of the seed page, the candidate page, and the seed page/candidate page pair; a ratio of a conversions rate for the seed page, a conversion rate for the candidate page, and a conversion rate for the seed page/candidate page pair; and a ratio of a conversion rate for the seed page/candidate page pair over a conversation rate for the seed page and a ratio of a conversion rate for the seed page/candidate page pair over a conversion rate for the candidate page.
14 . The system of claim 11 , further comprising categorizing the features based on demographic categories.
15 . The system of claim 11 , wherein the processing the features to generate a probability that a user will interact with the candidate content item comprises:
applying a machine learning technique based on the features.
16 . A non-transitory computer-readable storage medium including instructions that, when executed by at least one processor of a computing system, cause the computing system to perform a method comprising:
determining features based on online user behavior regarding a seed content item and a candidate content item that may be presented in response to an indication of approval by a user regarding the seed content item;
processing the features to generate a probability that the user will interact with the candidate content item; and
selecting the candidate content item for presentation to the user based on the probability that the user will interact with the candidate content item.
17 . The non-transitory computer-readable storage medium of claim 16 , wherein the seed content item is a seed page and the candidate content item is a candidate page.
18 . The non-transitory computer-readable storage medium of claim 17 , wherein the features include at least one of a number of conversions for each of the seed page, the candidate page, and a seed page/candidate page pair; a number of impressions for each of the seed page, the candidate page, and the seed page/candidate page pair; a ratio of a conversions rate for the seed page, a conversion rate for the candidate page, and a conversion rate for the seed page/candidate page pair; and a ratio of a conversion rate for the seed page/candidate page pair over a conversation rate for the seed page and a ratio of a conversion rate for the seed page/candidate page pair over a conversion rate for the candidate page.
19 . The non-transitory computer-readable storage medium of claim 16 , further comprising categorizing the features based on demographic categories.
20 . The non-transitory computer-readable storage medium of claim 16 , wherein the processing the features to generate a probability that a user will interact with the candidate content item comprises:
applying a machine learning technique based on the features.