Iterative Classifier Training on Online Social Networks
First Claim
1. A method comprising, by one or more computing devices:
- accessing a first set of objects associated with an online social network, each object being associated with one or more comments;
generating a second set of objects from the first set of objects by applying a first filtering criteria to the first set of objects;
scoring each object in the second set of objects based on the comments associated with each object;
generating a training set of objects from the second set of objects by selecting each object from the second set of objects having a score greater than a first threshold score, each object in the training set being associated with a first object-classification; and
determining an object-classifier algorithm for the first object-classification, the object-classifier algorithm being determined through an iterative training process performed one or more times, each iteration of the iterative training process comprising;
training an initial object-classifier algorithm based on the comments associated with the objects in the training set of objects;
accessing a third set of objects associated with the online social network;
classifying, using the initial object-classifier algorithm, each object in the third set of objects based on an analysis of the comments associated with each object, one or more of the objects in the third set of objects being classified with the first object-classification;
training a revised object-classifier algorithm based on the comments associated with the objects in the third set of objects having the first object-classification;
accessing a fourth set of objects associated with the online social network, the fourth set of objects being generated by applying a second filtering criteria to a fifth set of objects associated with the online social network;
classifying, using the revised object-classifier algorithm, each object in the fourth set of objects based on an analysis of the comments associated with each object, one or more objects in the fourth set of objects being classified with the first object-classification; and
generating a sixth set of objects from the fourth set of objects by selecting each object from the fourth set of objects having a score greater than a second threshold score, each object in the sixth set of objects being associated with the first object-classification, wherein the sixth set of objects is to be used as the training set in a next iteration of the iterative training process.
2 Assignments
0 Petitions
Accused Products
Abstract
In one embodiment, a method includes accessing a first set of objects associated with an online social network, each object being associated with one or more comments. The method also includes generating a second set of objects from the first set of objects by applying a first filtering criteria to the first set of objects and scoring each object in the second set of objects based on the comments associated with each object. The method further includes generating a training set of objects from the second set of objects by selecting each object from the second set of objects having a score greater than a first threshold score, each object in the training set being associated with a first object-classification. The method further includes determining an object-classifier algorithm for the first object-classification, the object-classifier algorithm being determined through an iterative training process performed one or more times.
-
Citations
20 Claims
-
1. A method comprising, by one or more computing devices:
-
accessing a first set of objects associated with an online social network, each object being associated with one or more comments; generating a second set of objects from the first set of objects by applying a first filtering criteria to the first set of objects; scoring each object in the second set of objects based on the comments associated with each object; generating a training set of objects from the second set of objects by selecting each object from the second set of objects having a score greater than a first threshold score, each object in the training set being associated with a first object-classification; and determining an object-classifier algorithm for the first object-classification, the object-classifier algorithm being determined through an iterative training process performed one or more times, each iteration of the iterative training process comprising; training an initial object-classifier algorithm based on the comments associated with the objects in the training set of objects; accessing a third set of objects associated with the online social network; classifying, using the initial object-classifier algorithm, each object in the third set of objects based on an analysis of the comments associated with each object, one or more of the objects in the third set of objects being classified with the first object-classification; training a revised object-classifier algorithm based on the comments associated with the objects in the third set of objects having the first object-classification; accessing a fourth set of objects associated with the online social network, the fourth set of objects being generated by applying a second filtering criteria to a fifth set of objects associated with the online social network; classifying, using the revised object-classifier algorithm, each object in the fourth set of objects based on an analysis of the comments associated with each object, one or more objects in the fourth set of objects being classified with the first object-classification; and generating a sixth set of objects from the fourth set of objects by selecting each object from the fourth set of objects having a score greater than a second threshold score, each object in the sixth set of objects being associated with the first object-classification, wherein the sixth set of objects is to be used as the training set in a next iteration of the iterative training process. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18)
-
-
19. One or more computer-readable non-transitory storage media embodying software that is operable when executed to:
-
access a first set of objects associated with an online social network, each object being associated with one or more comments; generate a second set of objects from the first set of objects by applying a first filtering criteria to the first set of objects; score each object in the second set of objects based on the comments associated with each object; generate a training set of objects from the second set of objects by selecting each object from the second set of objects having a score greater than a first threshold score, each object in the training set being associated with a first object-classification; and determine an object-classifier algorithm for the first object-classification, the object-classifier algorithm being determined through an iterative training process performed one or more times, each iteration of the iterative training process comprising; train an initial object-classifier algorithm based on the comments associated with the objects in the training set of objects; access a third set of objects associated with the online social network; classify, using the initial object-classifier algorithm, each object in the third set of objects based on an analysis of the comments associated with each object, one or more of the objects in the third set of objects being classified with the first object-classification; train a revised object-classifier algorithm based on the comments associated with the objects in the third set of objects having the first object-classification; access a fourth set of objects associated with the online social network, the fourth set of objects being generated by applying a second filtering criteria to a fifth set of objects associated with the online social network; classify, using the revised object-classifier algorithm, each object in the fourth set of objects based on an analysis of the comments associated with each object, one or more objects in the fourth set of objects being classified with the first object-classification; and generate a sixth set of objects from the fourth set of objects by selecting each object from the fourth set of objects having a score greater than a second threshold score, each object in the sixth set of objects being associated with the first object-classification, wherein the sixth set of objects is to be used as the training set in a next iteration of the iterative training process.
-
-
20. A system comprising:
- one or more processors; and
a memory coupled to the processors comprising instructions executable by the processors, the processors operable when executing the instructions to;access a first set of objects associated with an online social network, each object being associated with one or more comments; generate a second set of objects from the first set of objects by applying a first filtering criteria to the first set of objects; score each object in the second set of objects based on the comments associated with each object; generate a training set of objects from the second set of objects by selecting each object from the second set of objects having a score greater than a first threshold score, each object in the training set being associated with a first object-classification; and determine an object-classifier algorithm for the first object-classification, the object-classifier algorithm being determined through an iterative training process performed one or more times, each iteration of the iterative training process comprising; train an initial object-classifier algorithm based on the comments associated with the objects in the training set of objects; access a third set of objects associated with the online social network; classify, using the initial object-classifier algorithm, each object in the third set of objects based on an analysis of the comments associated with each object, one or more of the objects in the third set of objects being classified with the first object-classification; train a revised object-classifier algorithm based on the comments associated with the objects in the third set of objects having the first object-classification; access a fourth set of objects associated with the online social network, the fourth set of objects being generated by applying a second filtering criteria to a fifth set of objects associated with the online social network; classify, using the revised object-classifier algorithm, each object in the fourth set of objects based on an analysis of the comments associated with each object, one or more objects in the fourth set of objects being classified with the first object-classification; and generate a sixth set of objects from the fourth set of objects by selecting each object from the fourth set of objects having a score greater than a second threshold score, each object in the sixth set of objects being associated with the first object-classification, wherein the sixth set of objects is to be used as the training set in a next iteration of the iterative training process.
- one or more processors; and
Specification