Facebook, Inc.

Jeffrey Scott Dunn

Cameron Marlow

Venkataramanan Nandagopal

Rafael Linden Sagula

Robert William Cathcart

Joel F Beales

Siddharth Kar

Alisson Gusatti Azzolini

Stuart Michael Bowers

Hussein Mohamed Hassan Mehanna

Donghang Guo

James Robert Paton

Xiaowen Xie

Aleksandr Sidorov

Pamela Shen Vagata

Rodrigo Bouchardet Farnham

Philip Lewis Bohannon

Mukund Narasimhan

Eric Sun

Chad E Little

Yuankai Ge

Jia Li

Scott Kenneth Yost

Hangjin Zhang

Shai Duvdevani

Le Zhang

Jonathan D Chang

Eun Chang Lee

Eytan Bakshy

Jonathan Michael Wiepert

Luke Reynold Styles

David Ross Harrington

Lifei Huang

Alexander Sebastiaan Mols

Phyllipe Cesar Ramos De Almeida Medeiros

Lawrence Lomax

Rajesh Yengisetty

Alexander Juarez

Quang Minh Tuan Nguyen

Zsolt Dollenstein

Waclaw Jan Banasik

Cameron Alexander Marlow

Venkataramanan Iyer Nandagopal

A computer-implemented method for identifying and tracking application performance incidents may include (1) receiving, by an incident tracking system, data representative of a time series, the time series including a time-ordered plurality of values of a performance metric associated with a program, (2) identifying, by the incident tracking system, a discontinuity in the time series, (3) associating, by the incident tracking system, the identified discontinuity in the time series with a change in source code associated with the program, and (4) executing, by the incident tracking system, an automated action based on the association of the identified discontinuity with the change in the source code. Various other methods, systems, and computer-readable media are also disclosed.

Systems and methods for identifying and tracking application performance incidents

Some embodiments include a workflow authoring tool that accesses a text string representation of a workflow and a text string representation of at least a data processing operator type. The workflow authoring tool enables definition of one or more data processing operator types that can be referenced in defining the machine learning workflow. When scheduling a workflow, the text string representation of the workflow can be parsed and traversed to generate an interdependency graph of one or more data processing operators. The text string representation of the data processing operator type can identify operator attributes associated with the data processing operator type.

Machine learning system flow authoring tool

A computer-implemented method for distributed management of computing resources may include (i) performing, by a computing device, an initial configuration of one or more computing resources connected to a network, (ii) detecting a request for a computing resource from a client daemon, (iii) based on the request, initializing a computing environment on the computing resource, (iv) maintaining an active state of the computing resource for a usage session by a client device, (v) detecting, from the client daemon, a notification of completion of the usage session, and (vi) in response to the notification of completion, reverting the computing resource to an initial state. Various other methods, systems, and computer-readable media are also disclosed.

Systems and methods for distributed management of computing resources

A computer-implemented method for improving comparative performance test results of mobile applications may include (1) determining an optimum testing configuration for a mobile computing device, (2) directing the mobile computing device to (a) execute a comparative performance test, (b) operate in accordance with the determined optimum testing configuration during the execution of the comparative performance test, and (c) write data generated during the execution of the comparative performance test to a random-access memory (RAM) drive of the mobile computing device, (3) recording a network response directed to the mobile computing device, (4) detecting a subsequent network request sent by the mobile computing device, (5) sending the recorded network response to the mobile computing device in response to detecting the subsequent network request, and (6) tracking a control performance value and a modified performance value during the comparative performance test. Various other methods and systems are also disclosed.

Systems and methods for improving comparative performance test results of mobile applications

Some embodiments include an experiment management interface for a machine learning system. The experiment management interface can manage one or more workflow runs related to building or testing machine learning models. The experiment management interface can receive an experiment initialization command to create a new experiment associated with a new workflow. A workflow can be represented by an interdependency graph of one or more data processing operators. The experiment management interface enables definition of the new workflow from scratch or by cloning and modifying an existing workflow. The workflow can define a summary format for its inputs and outputs. In some embodiments, the experiment management interface can automatically generate a comparative visualization at the conclusion of running the new workflow based on an input schema or an output schema of the new workflow.

Machine learning system interface

Some embodiments include a method of machine learner workflow processing. For example, a workflow execution engine can receive an interdependency graph of operator instances for a workflow run. The operator instances can be associated with one or more operator types. The workflow execution engine can assign one or more computing environments from a candidate pool to execute the operator instances based on the interdependency graph. The workflow execution engine can generate a schedule plan of one or more execution requests associated with the operator instances. The workflow execution engine can distribute code packages associated the operator instances to the assigned computing environments. The workflow execution engine can maintain a memoization repository to cache one or more outputs of the operator instances upon completion of the execution requests.

Machine learning system flow processing

A system and method that automatically detects that a regression incident has occurred, and identifies the particular revision(s) of source code that introduced the regression incident is disclosed herein. The system and method execute one or more simulations of each revision in a window of revisions of the source code to compute values for one or more metrics. The metric values are then used to compute regression confidence values as well as a regression confidence interval&#x2014;to then detect a regression incident and identify the revision that caused the regression incident. The system and method can also identify subsequent revision(s) that sufficiently fix regression incident(s) caused by a revision such that a software developer can be informed of these fixing revisions.

Incident tracker

There are provided means for implementing an interface to view and explore socially relevant concepts of an entity graph including, for example, means of a social network system to perform operations including retrieving contextually relevant data for a plurality of concepts within an entity graph of the social network system; retrieving socially relevant data for a user's node within a social graph of the social network system; identifying intersects between the plurality of concepts within the entity graph and the social relevant data for the user's node within the social graph; selecting one of the plurality of concepts within the entity graph based on the intersects identified; and displaying the one of the plurality of concepts within the entity graph at a user interface associated with the user's node.

Systems, methods, and apparatuses for implementing an interface to view and explore socially relevant concepts of an entity graph

Embodiments improve the ability of users of a social networking system to search for information that is likely to be relevant to them by learning and/or applying a search context associated with selector components used to search for objects of the social networking system. The search context is specific to the use of an individual selector and thus need not be as general as the context of an entire page or set of pages in which selectors can be embedded. The social networking system may learn the context of a selector by monitoring user selections from prior search results performed using the selector.

Context-based object retrieval in a social networking system

A social networking system leverages user's social information to evaluate content submitted for inclusion in objects. If the evaluated submission is accepted, the submission is added to the content of an object. Accepted submissions are also used to predict associations between metadata and objects. Metadata is used to predict which objects will match user searches for information. The social networking system also provides a user interface configured to prompt users to submit information to objects. When a user completes a submission to an object, the user is provided with other options for groups of objects to contribute to. The objects offered are chosen to increase the likelihood that the user will choose to provide submissions to one of the provided objects.

Metadata prediction of objects in a social networking system using crowd sourcing

The disclosure is directed to a benchmarking system for measuring performance of a client-side application, e.g., a web browser, in processing an application, e.g., rendering a web page of a social networking application. The benchmarking process is executed in multiple modes, e.g., a record mode and a replay mode. In the record mode, the benchmarking system warms up a proxy server by storing request-response pairs between a client device and an app server in a cache of the proxy server. In the replay mode, the benchmarking system replays the requests to obtain the responses from the cache of the proxy server and records various metrics that indicate a performance of the client-side application in processing the responses, e.g., rendering the web page.

Methods and systems for benchmarking web browser performance

In one embodiment, a method includes generating an executable version of a decision tree by compiling source code into executable code and verifying the executable code by comparing a result of executing the executable code with a result of evaluating the decision tree in interpreted mode. The method further includes replacing the decision tree evaluated in the interpreted mode with the executable code if the executable code is verified or discarding the executable code otherwise.

Decision tree ensemble compilation

In one embodiment, a decision tree is evaluated in interpreted mode while statistics are collected. The decision tree is then represented as source code, and each decision in the decision tree is annotated with instructions determined based on the collected statistics. The source code is compiled into machine code, and the machine code is optimized based on the instructions annotating each decision in the decision tree.

A social networking system selects a set of social endorsements for display within or in conjunction with an advertisement. Candidate social endorsements are identified in response to receiving a request for social endorsements information, each associated with an amount of display space, an affinity with a viewing user, and one or more social networking system objects. The amount of space available to display social endorsement information is determined, and sets of candidate social endorsements are generated to fit within the amount of space available to display social endorsements. The sets of candidate social endorsements are ranked, for instance based on the affinities associated with the candidate social endorsements within each set of candidate social endorsements. A set of candidate social endorsements is selected based on the ranking, and is provided for display within or in conjunction with an advertisement.

Adaptive selection of social context for display to a viewing user

Growing community of inventors

Seattle, WA, United States of America

Jeffrey Scott Dunn