Abstract: There is described a system and computer-implemented method for providing a recommendation based on a sparse pattern of data. An exemplary method comprises determining a likelihood that an item for which no user preference data is available will be preferred. The exemplary method also comprises determining a likelihood that an item for which user preference data is available for users other than a particular user will be preferred based on the likelihood that the item for which no user preference data is available will be preferred. The exemplary method additionally comprises predicting that an item for which no user preference data relative to the particular user is available will be preferred if the likelihood that the particular user will prefer the item exceeds a certain level.

Abstract: There is described a system and computer-implemented method for providing a recommendation based on a sparse pattern of data. An exemplary method comprises determining a likelihood that an item for which no user preference data is available will be preferred. The exemplary method also comprises determining a likelihood that an item for which user preference data is available for users other than a particular user will be preferred based on the likelihood that the item for which no user preference data is available will be preferred. The exemplary method additionally comprises predicting that an item for which no user preference data relative to the particular user is available will be preferred if the likelihood that the particular user will prefer the item exceeds a certain level.

Abstract: There is described a system and computer-implemented method for providing a recommendation based on a sparse pattern of data. An exemplary method comprises determining a likelihood that an item for which no user preference data is available will be preferred. The exemplary method also comprises determining a likelihood that an item for which user preference data is available for users other than a particular user will be preferred based on the likelihood that the item for which no user preference data is available will be preferred. The exemplary method additionally comprises predicting that an item for which no user preference data relative to the particular user is available will be preferred if the likelihood that the particular user will prefer the item exceeds a certain level.

Abstract: For each first entity of a subset of a number of first entities, an expected improvement of a predictive performance of a collaborative filtering model if additional ratings of the first entity in relation to a plurality of second entities were obtained is estimated. Particular first entities from the subset of the first entities of which to obtain the additional ratings in relation to the second entities are selected based at least on the expected improvements that have been determined. The additional ratings of the particular first entities in relation to the second entities are obtained.

Abstract: Developing a population category hierarchy can include providing a candidate category hierarchy, including a number of candidate categories, and a mapping between a number of reference pages and the number of candidate categories, including a number of mapped reference pages (143). Population usage data of the number of mapped reference pages can be obtained and used to determine a population traffic metric for each of the number of candidate categories (147). A number of population categories can be generated by using the population traffic metric of each of the number of candidate categories (149); and, a population category hierarchy can be produced including the number of population categories (151).

Abstract: One embodiment is a method that receives a seed Uniform Resource Locator (URL) that represents a category for website classification. Clickstream data generated from the seed URL and additional URLs are analyzed to determine whether the additional URLs belong to the category. The method selects one or more of the additional URLs to represent the category.

Abstract: An exemplary embodiment of the present invention provides a method of processing Web activity data. The method includes obtaining a database of Website organizational data. The method also includes generating a data structure from the database of Website organizational data comprising an Item identifier and a Website category corresponding to the item identifier. The method also includes generating a reduced-rank classification structure from the data structure, the reduced-rank classification structure including a category grouping corresponding to one or more of the Website categories.

Abstract: A method performed by a processing system includes receiving a recommendation from a source user in response to performing an action corresponding to an action context of the recommendation, determining whether the source user appears in social network information of a target user, and distinguishing a presentation of the recommendation to the target user in response to the source user appearing in the social network information of the target user.

Abstract: For each web page visited, a path is determined through a hierarchy of categories. The hierarchy of categories has levels from a most abstract level to a most concrete level. For each microblog entry of a microblog, a path is determined through the hierarchy of categories. Each microblog entry for which the path is similar to the path for at least one web page is determined as a selected microblog entry.

Abstract: In the present disclosure, methods and apparatuses are disclosed that enable a device to determine whether a contact is in a shared environment based on an audio sample of a voice call. More specifically, an audio sample of a voice call is generated. A controller then determines whether a contact is in an environment of the mobile device based on the audio sample.

Abstract: A collection of labeled training cases is received, where each of the labeled training cases has at least one original feature and a label with respect to at least one class. Non-linear transformation of values of the original feature in the training cases is applied to produce transformed feature values that are more linearly related to the class than the original feature values. The non-linear transformation is based on computing probabilities of the training cases that are positive with respect to the at least one class. The transformed feature values are used to train a classifier.

Abstract: Systems, methods, and machine readable and executable instructions are provided for collaborative filtering. Collaborative filtering includes representing users and objects by rows and columns in a binary ratings matrix having a particular dimensional space. Unknown values in the binary ratings matrix are weighted with a weight matrix having the particular dimensional space. The binary ratings matrix and the weight matrix are hashed into a lower dimensional space by one of row and column. The hashed binary ratings matrix and the hashed weight matrix are low-rank approximated by alternating least squares. A result of the low-rank approximation for the one of row and column is updated using the binary ratings matrix and the weight matrix. A recommendation of one of the objects can be generated for one of the users based on the updated result.

Abstract: Systems, methods, and machine readable and executable instructions are provided for collaborative filtering. Collaborative filtering includes representing users and objects by rows and columns in an ordinal ratings matrix having a particular dimensional space. Values in the ordinal ratings matrix are weighted with a weight matrix having the particular dimensional space. The weight matrix is hashed into a lower dimensional space by one of row and column by multiplying a projection matrix by the weight matrix. The ordinal ratings matrix is hashed into a lower dimensional space by multiplying the projection matrix by an element-wise product of the weight matrix and the ordinal ratings matrix to form a reduced ratings matrix, and element-wise dividing the reduced ratings matrix by the hashed weight matrix. The hashed ordinal ratings matrix and the hashed weight matrix are low-rank approximated by alternating least squares.

Abstract: Systems (490), methods (100, 200), and computer-readable and executable instructions (324, 424) are provided for estimating costs of behavioral targeting. Estimating costs of behavioral targeting can include scoring a topic with a behavioral targeting model (101, 201). Estimating costs of behavioral targeting can also include obtaining a plurality of data items including geographic location information (102, 202). Estimating costs of behavioral targeting can also include detecting (104, 204) and scoring (209) a sentiment from filtered data items regarding a topic within a region (104, 204). Estimating costs of behavioral targeting can include computing a penalty score for the topic in the region in response to the scored sentiment exceeding a threshold (213), (106, 206). Estimating costs of behavioral targeting can include adjusting the topic score in the region according to the penalty score (108, 208).

Abstract: Developing a population category hierarchy can include providing a candidate category hierarchy, including a number of candidate categories, and a mapping between a number of reference pages and the number of candidate categories, including a number of mapped reference pages (143). Population usage data of the number of mapped reference pages can be obtained and used to determine a population traffic metric for each of the number of candidate categories (147). A number of population categories can be generated by using the population traffic metric of each of the number of candidate categories (149); and, a population category hierarchy can be produced including the number of population categories (151).

Abstract: Improving data clustering stability. A computer accesses a first plurality of cluster groups comprising data. The computer then applies a clustering method to the first plurality of cluster groups while adjusting said first plurality of cluster groups to be in higher agreement between themselves, thereby generating a second plurality of cluster groups that is in higher agreement between themselves than the first plurality of cluster groups. The second plurality of cluster groups corresponds to the first plurality of cluster groups.

Abstract: A method performed by a processing system includes receiving a recommendation from a source user in response to performing an action corresponding to an action context of the recommendation, determining whether the source user appears in social network information of a target user, and distinguishing a presentation of the recommendation to the target user in response to the source user appearing in the social network information of the target user.

Abstract: Systems, methods, and computer-readable and executable instructions are provided for presenting streaming media for an event. Presenting streaming media for an event can include receiving an outgoing message. Presenting streaming media for an event can also include evaluating the outgoing message to determine if the message is related to the event. Furthermore, presenting streaming media for an event can include taking an action with respect to presenting streaming media based on the relation to the event.

Abstract: A computer-implemented method comprises running, by a processor, a plurality of classifiers on a web page to obtain one or more keywords. The method further comprises selecting, by the processor, advertisements based on the one or more keywords.

Abstract: Systems, methods, and machine readable and executable instructions are provided for collaborative filtering. Collaborative filtering includes representing users and objects by rows and columns in an ordinal ratings matrix having a particular dimensional space. Values in the ordinal ratings matrix are weighted with a weight matrix having the particular dimensional space. The weight matrix is hashed into a lower dimensional space by one of row and column by multiplying a projection matrix by the weight matrix. The ordinal ratings matrix is hashed into a lower dimensional space by multiplying the projection matrix by an element-wise product of the weight matrix and the ordinal ratings matrix to form a reduced ratings matrix, and element-wise dividing the reduced ratings matrix by the hashed weight matrix. The hashed ordinal ratings matrix and the hashed weight matrix are low-rank approximated by alternating least squares.