Meta Platforms, Inc.

Aaron Bryan Adcock

Jan Kodovsky

Justin Moore

Yaniv Shmueli

Jan Kalis

Danielle Rothermel

Zhong Wang

Jonathan Haski

Aileen Chen

Abhishek Parthasarathy

Myle Arif Ott

Peng-Jen Chen

Konrad Komorowski

Wenbo Yuan

Junfei Wang

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

An online system maintains various labels identifying physical locations, allowing users to identify a physical location to the online system by specifying the label identifying the physical location. When a user does not identify a label corresponding to location coordinates provided to the online system by a client device, the online system selects a maintained label and prompts the user to indicate whether the user visited the physical location identified by the selected label. To select the label, the online system accounts for likelihoods of the received location coordinates being identified by various labels and values of the online system of identifying whether different labels identify received location coordinates. The online system may also account for other prompts presented to the user or to other users when determining whether to prompt the user to indicate whether the user visited the physical location identified by the selected label.

Determining correlations between locations associated with a label and physical locations based on information received from users providing physical locations to an online system

In one embodiment, a method includes receiving from a client system of a user, background signal-information identifying one or more first wireless signals within wireless communication range of the client system; accessing a place-entity database, wherein the place-entity database comprises information indicating that a first place-entity corresponds to one or more second wireless signals; determining that the client system is located at a geographic location associated with the first place-entity based on determining that the one or more first wireless signals match the one or more second wireless signals and further based on the information indicating that the first place-entity corresponds to the one or more second wireless signals; and sending, to the client system, information associated with the first place-entity automatically without a query from the user of the client system, wherein the query is related to the first place-entity.

Location prediction using wireless signals on online social networks

In one embodiment, a method includes sending information associated with multiple candidate place-entities to a user, receiving a selection of one candidate place-entity by the user, and updating the rating of the selected candidate place-entity on an online social network. In particular, the sending is based on calculated probabilities the candidate place-entities are associated with a particular geographic location.

Location-based place determination using online social networks

In one embodiment, a method includes receiving, from a first software application of a client system associated with a user of an online social network, background signal-information identifying one or more first wireless signals; storing the signal-information and a client identifier for the client system in a signal-information database; receiving, from a second software application of the client system via a places-API of the online social network, a places-API call indicating that the client system is located at a geographic location corresponding to a first place-entity; recording the places-API call in an API-call log, wherein the API-call log records the first place-entity and the client identifier; determining a correlation between the signal-information and the first place-entity; and updating a place-entity database to indicate that the first place-entity corresponds to the one or more first wireless signals identified by the signal-information.

In one embodiment, one or more computing devices of an online social network, receives, from a mobile-client system, geographic-location information associated with the mobile-client system. The one or more computing devices identify multiple candidate place-entities associated with the online social network that correspond to the geographic-location information, where each candidate place-entity is associated with a particular geographic location. The method also includes determining, for each candidate place-entity, a confidence score based on the geographic-location information associated with the mobile-client system, where the confidence score represents a probability that the first user is located at the candidate place-entity. A location-probability distribution associated with the candidate place-entity can also be used to calculate the confidence score. The one or more computing devices send, to the mobile-client system, information associated with one or more of the candidate place-entities based on their respective confidence scores.

In one embodiment, a method includes, by one or more computing devices of an online social network, receiving, from a mobile-client system of a first user of the online social network, geographic-location information associated with the mobile-client system. The method further includes identifying multiple candidate place-entities associated with the online social network that correspond to the geographic-location information, where each candidate place-entity is associated with a particular geographic location. The method also includes determining, for each candidate place-entity, a confidence score based on the geographic-location information associated with the mobile-client system and a location-probability distribution associated with the candidate place-entity, where the confidence score represents a probability that the first user is located at the candidate place-entity. The method also includes sending, to the mobile-client system of the first user, information associated with one or more of the candidate place-entities having a confidence score above a threshold confidence score.

Growing community of inventors

New York, NY, United States of America

Aaron Bryan Adcock