Facebook, Inc.

Other

Michael Dudley Johnson

Ryan Patterson

Erick Tseng

William Joseph Flynn, Iii

Tao Xie

Joshua Richard Williams

Hsi Wang

Eyal Michael Sharon

Xiaoliang Wei

Herry Sutanto

Mukund Narasimhan

Yan Yu

Breno Pompeu Roberto

Itamar Rosenn

Alex Bain

Weizhe Shi

Todd Krabach

Hu Li

Yohann Richard

Lior Berry

Hsueh-Fang Cheng

Mathieu Benjamin Tozer

Jackson Gabbard

Eric Ruth

Breno Pompeu Robert

The present disclosure provides methods and systems for generating linguistic rules. The system may comprise: an electronic display with a graphical user interface comprising: (i) one or more interactive elements for receiving an user input indicating one or more edits to a rule, and (ii) a result visualization region for dynamically displaying a result of the rule in response to receiving the one or more edits, wherein the result of the rule comprises an indicator indicating the validity of the rule; and one or more computer processors that are programmed to: (i) generate the result of the rule by processing the rule with the one or more edits against a set of examples; and (ii) configure the graphical user interface to display the result in a user-selected format.

Methods and systems for generating linguistic rules

The present disclosure provides methods and systems for generating a rule for machine translation. The method comprises receiving a user input indicating a change of a rule; retrieving a prior version of the rule from a storage unit upon receiving the user input; identifying one or more metrics from a plurality of metrics based at least in part on the prior version of the rule; computing, with aid of one or more processors, a value for each of the one or more metrics by evaluating each metric against a set of examples, wherein the value represents the impact of the change of the rule; and comparing the value to a pre-determined threshold to determine whether the change of the rule is acceptable.

Methods and systems for scalable machine translation

In one embodiment, one or more main processors of a device detect an event corresponding to scrolling a structured document presented in a scrollable graphical user interface of an application executed by the one or more main processors. The main processors, in response to the event, determine a first set of elements and a second set of elements of the structured document; render the first set of elements of the structured document for presentation in the graphical user interface; and cause one or more graphics processors of the device to render the second set of elements for presentation in the graphical user interface.

Smooth scrolling of a structured document presented in a graphical user interface with bounded memory consumption

In one embodiment, a method includes, sending, to a client device of a user of an online social network for presentation to the user, one or more references corresponding to one or more content items of the online social network, respectively. Each reference is presented with one or more call-to-action elements. Each call-to-action element is a user-interface element that prompts an action from the user to initiate a particular function or process related to the content item via the online social network. The call-to-action elements for each content item are presented in ranked order based on a suitability of the call-to-action element to the content item. The method includes receiving a selection of a call-to-action element presented with a reference corresponding to a particular content item. The method includes executing the function or process related to the content item on the online social network prompted by the selected call-to-action element.

Context-based selection of calls-to-action associated with search results

In one embodiment, a method includes accessing a base user experience (UX) including a user interface (UI) corresponding to one or more features of a computing system; determining one or more user categories based at least in part on previous interactions with the UI by a number of users; classifying, using a machine-learning algorithm, a particular user into one or more of the user categories based at least in part on current interactions of the particular user with the UI; and determining, relative to the base UX, one or more modifications to the UI based at least in part on the classification of the particular user into one or more of the user categories. The modifications to the UI modify one or more features of the UX. The method also includes applying the modifications to the UI; and providing the UI as modified for display to the particular user.

User experience/user interface based on interaction history

In one embodiment, a method includes compiling one or more references to content items based on information known about a user stored in one or more object stores of a social-networking system. Each reference is associated with one or more call-to-action elements applicable to the content item associated with the reference. The call-to-action elements are selectable by the user to initiate the execution of an action modifying the related content item on the social-networking system. Modifying the related content item may include modifying metadata associated with the content item in one or more object stores of the social-networking system. The method includes sending, to a client device of the user after the user accesses a tool provided by the social-networking system, the references with their associated call-to-action elements for presentation to the user. The call-to-action elements are presented to the user in proximity to their associated references.

Suggesting search results to users before receiving any search query from the users

A user interface includes a tactile interface on a client device that allows users to interact with social networking system content in a manner that is optimized for touch screens and/or devices with limited display area. Stories associated with social networking system users and times are displayed in a scrollable list along with a timeline scrubber identifying different time periods. A position indicator is displayed proximate to the timeline scrubber to identify the time period including one or more times associated with displayed stories. Responsive to received input, details about time intervals included in a time period are displayed or the timeline scrubber appearance is modified to distinguish certain time periods.

Dynamically scaled navigation system for social network data

In one embodiment, a computing system accesses current usage information of a mobile device associated with a user. The computing system then determine a future user state of the user based on multiple predictor functions. The multiple predictor functions calculate a probability of the future user state based on multiple weights and multiple decay factors. The multiple decay factors are determined from past user states. The computing system further sends instructions to the mobile device for adapting the operation of the mobile device corresponding to the determined future user state.

Predicting the future state of a mobile device user

In one embodiment, a first computing server receives a message from a client computing device. The message identifies an image to be displayed on the client computing device. The message is generated by a second computing server and received by the client computing device in response to a request sent by the client computing device to the second computing server for the identified image. The first computing server determines one or more adjustment parameters based on the display capabilities of the client computing device. The first computing server creates an adjusted version of the identified image based on the adjustment parameters, and sends the adjusted version of the identified image to the client computing device for display.

Photo selection for mobile devices

In one embodiment, a method includes, by one or more server computing devices, receiving state data of a client computing device. The state data includes event data indicating events generated by or occurring at the client computing device and context data associated with the event data. The context data indicates device states of the client computing device that each coincide with one or more of the events and indicate a context of the one or more of the events. The method also includes, by one or more server computing devices, ordering the events and the device states in the event and context data into a state-data-review structure and analyzing the state-data-review structure to generate one or more recommendations on operation of the client computing device.

Device state capture and analysis

In one embodiment, a method includes deploying an application to a plurality of client systems associated with a plurality of users, respectively, wherein the application is installed on each of the client systems, and wherein one or more treatments can be remotely activated on the application, receiving, from a first client system associated with a first user of the plurality of users, a request from the application installed on the first client system to access the application server, determining whether the first user belongs in a treatment group associated with a first treatment, remotely activating the first treatment on the first client system associated with the first user if the first user belongs in the treatment group, and synchronizing activation of the first treatment in response to subsequent access by the first client system.

Native application testing

In one embodiment, a social-networking system may compile a set of search results based on information known about a user stored by the social-networking system, the search results being compiled before the user inputs any search query or portion thereof, each search result being associated with one or more call-to-action elements applicable to the search result, each call-to-action element prompting an action from the user related to the search result via the social-networking system, and send the set of search results with the call-to-action elements for presentation to the user, wherein the call-to-action elements are presented to the user in proximity to their associated search results.

In one embodiment, a method includes receiving a search query for content items of the online social network, identifying content items of the online social network that match the search query, and sending, in response to the search query, search results corresponding to the identified content items, where each search result is presented with one or more call-to-action elements applicable to the content item, and where the call-to-action elements are presented in ranked order based on a suitability of the call-to-action element to the content item.

In one embodiment, a computing system receives an indication of current mobile device usage by a user. The user is also associated with one or more past user states that are accessible to the computing system. The computing system selects a future user state of the user based on a calculated probability of a possible future user state. The possible future user state is calculated based on the current mobile-device usage and the past user states. For each past user states, a weight based on a time delay factor is used. Based on the selected future user state, the operation of a mobile device of the user is adapted at the future time corresponding to the future user state.

Growing community of inventors

San Mateo, CA, United States of America

Michael Dudley Johnson