International Business Machines Corporation

Facebook, Inc.

Pinterest, Inc.

Aditya Pal

Fei Wang

Charles Joseph Rosenberg

Jalal U Mahmud

Michelle Xue Zhou

Deepayan Chakrabarti

Jurij Leskovec

Chantat Eksombatchai

Bo Zhao

Yitong Zhou

Wei Zhang

Fei C Wang

Anitha Kannan

Fei Huang

Liang Gou

Amaç Herdagdelen

Huahai Yang

Sourav Chatterji

Stanislav Funiak

Daniel Matthew Merl

Karthik Subbian

Mengdie Hu

Seyoung Park

Sumit Taank

Systems and methods for recommending content to an online service subscriber are presented. For each subscriber, content items that were the subject of the subscriber's prior interactions are projected, via associated embedding vectors, into a content item embedding space. The content items, via their projections into the content item embedding space, are clustered to form a plurality of interest clusters for the subscriber. A representative embedding vector is determined for each interest cluster, and a plurality of these embedding vectors are stored as the representative embedding vectors for the subscriber. The online service, in response to a request for recommended content for a subscriber, selects a first representative embedding vector associated with the subscriber and identifies a new content item from a corpus of content items according to a similarity measure between the first representative embedding vector and an embedding vector associated with the new content item.

Recommending content to subscribers

In one embodiment, a system may access a graph data structure that includes nodes and connections between the nodes. Each node may be associated with a user; each connection between two nodes may represent a relationship between the associated users; and each node may be either labeled or unlabeled with respect to a label type. For each labeled node, a label of the label type of that labeled node may be propagated to other nodes through the connections. For each node, the system may store a label distribution information associated with the label type based on the propagated labels reaching the node. The system may train a machine-learning model using the labels and the label distribution information of a set of the labeled nodes. A predicted label for each unlabeled node may be generated using the model and the label distribution information of the unlabeled node.

Semi-supervised learning via deep label propagation

Systems, methods, and non-transitory computer-readable media can determine one or more respective topics of interest for at least some users of a social networking system. At least some of the topics can be propagated to at least a first user, wherein the propagated topics were determined to be of interest to users that follow the first user in the social networking system. At least one topic from the propagated topics for which the first user is a topical authority is determined.

Systems and methods for recommendation of topical authorities

In a method for determining group attributes and matching tasks to a group, a plurality of individual attributes for members of a first group of a plurality of groups are determined, wherein each individual attribute has a type. Parameters of a first distribution of at least one type of individual attribute across members of the first group are estimated. Group attributes of the first group are determined based, at least in part, on the estimated parameters of the first distribution of at least one type of individual attribute. The determined group attributes of the first group are stored in a repository, wherein the repository includes group attributes associated with each group of the plurality of groups. A task is received, wherein the task is associated with a specific group attribute and the task is matched to one group of the plurality of groups based on the specific group attribute.

Determining group attributes and matching tasks to a group

A computer-implemented method routes a current question to one or more of a plurality of online communities. A computer system can determine, for the current question presented by an asking user a plurality of question-to-question similarity values, a plurality of question-to-user similarity values and a plurality of question-to-community similarity values. The system can select one or more of the plurality of online communities based on the similarity values. The system can route the current question presented by the asking user to the selected one or more of the plurality of online communities.

Question routing for user communities

A system for identifying language(s) for content items is disclosed. The system can identify different languages for content item words segments by identifying segment languages that maximize a probability across the segments. The probability can be a combination of: an author's likelihood for the language identified for the first word; a combination of transition frequencies for selected languages identified for words, the transition frequencies indicating likelihoods that a transition occurred to the selected language from the previous word's language; and a combination of observation probabilities indicating, for a given word in the content item, a likelihood the given word is in the identified language. For an in-vocabulary word, the observation probabilities can be based on learned probability for that word. For an out-of-vocabulary word, the probability can be computed by breaking the word into overlapping n-grams and computing combined learned probabilities that each n-gram is in the given language.

Identifying multiple languages in a content item

Embodiments relate to relationship modeling and visualization from social media. One aspect includes determining a relationship type of a network-based relationship, between an individual and a network contact of the individual, from at least one social media data source. The relationship type is determined using a relationship model based on relationship types that include operational, personal, and business. Another aspect includes performing timeline based relationship strength segmentation using Group Lasso. The timeline based relationship strength segmentation specifies a past and current strength of the relationship. A further aspect includes predicting a future strength of the relationship using Extended Kalman Filter, and providing, through a visual interface, interactive visual analytics to view and monitor relationship states including the past, current, and future strengths over time.

Modeling and visualizing a dynamic interpersonal relationship from social media

Embodiments of the invention relate to assessing characteristics of a message and a message recipient. A trust model is established to take into account a set of trust antecedents, including characteristics of the messages and properties of the recipients, a set of action motivations, and their contribution to the action-based trust measurement. The assessment(s) is utilized to produce a tangible trust measurement that is employed to gauge the recipient's perception of credibility towards the received message.

Action based trust modeling

Growing community of inventors

San Jose, CA, United States of America

Aditya Pal