Facebook, Inc.

Meta Platforms, Inc.

Myle Arif Ott

Yaniv Shmueli

Jason Luis Briceno

Justin Moore

Aaron Bryan Adcock

Marian Gelu Olteanu

Peng-Jen Chen

Long Chen

Wenbo Yuan

Junfei Wang

Ivan Medvedev

In one embodiment, a method includes receiving, from a client system associated with a user of an online social network, data indicating that the user is located at a first geographic location at a first time; accessing a first embedding representing a first place-entity corresponding to the first geographic location; accessing multiple second embeddings representing multiple respective second place-entities each corresponding to a second geographic location; calculating, a similarity metric between the embedding representing the first place-entity and each of the embeddings representing the second place-entities; ranking each of the second place-entities based on their calculated similarity metrics; and sending, to the client system, information associated with one or more second geographic locations corresponding to one or more second place-entities having a ranking greater than a threshold ranking.

Prediction of next place visits on online social networks

In one embodiment, a method includes receiving from a client system a search query, identifying a first place-entity based on the search query, accessing a place-entities graph comprising a plurality of place-entity nodes, each place-entity node representing a particular place-entity associated with a particular geographic location, wherein the first place-entity is represented by a first place-entity node, accessing a redirection graph comprising the plurality of place-entity nodes and a plurality of place-entity clusters, each place-entity node in a place-entity cluster having a redirection edge connecting the place-entity node to a canonical place-entity node of the respective place-entity cluster, and sending a response to the search query, wherein if the first place-entity node is connected to a canonical place-entity node by a redirection edge within the redirection graph, the response comprises a reference to the canonical place-entity node, else the response comprises a reference to the first place-entity node.

Suppression and deduplication of place-entities on online social networks

In one embodiment, an online social network accesses a place-entity cluster comprising a number of place-entity nodes corresponding to a particular place-entity having a geographic location. One of the place-entity nodes is identified as an initial canonical place-entity cluster connected to the other place-entity nodes by redirection edges. A cluster score is calculated for each place-entity node in the cluster, and nodes having a cluster score above a threshold is identified. One of the identified place-entity nodes is selected as a replacement canonical place-entity node. If the replacement node is different from the initial canonical node, then the place-entity cluster is updated by adding or removing at least one place-entity node from the cluster based on their respective cluster scores.

Identification of real-best-pages on online social networks

In one embodiment, an online social network accesses a place-entities graph that includes a plurality of place-entity nodes, each place-entity node representing a particular place-entity associated with a particular geographic location. From one of a plurality of search clients associated with the online social network, a threshold duplication value and a threshold redirect value is received. The online social network identifies place-entity clusters of place-entity nodes within the place-entities graph. In each place-entity cluster, the place-entity nodes have duplication values with respect to a canonical place-entity node in the cluster that are above the threshold duplication value. The online social network generates a redirection graph for the one search client. The redirection graph includes the identified place-entity clusters, where for each cluster, a redirection edge is established between a place-entity node and the corresponding canonical place-entity node based on the duplication value and the threshold redirect value.

Growing community of inventors

New York, NY, United States of America

Myle Arif Ott