Training data quality for spam classification

    公开(公告)号:US11232369B1

    公开(公告)日:2022-01-25

    申请号:US15698797

    申请日:2017-09-08

    Applicant: Facebook, Inc.

    Abstract: In one embodiment, a method includes accessing posts in a social-networking system. Each of the posts is unlabeled with respect to whether the post is known to be spam. The method also includes determining a posting user who submitted the post to the social-networking system and a recipient user to whom the post is addressed. The method further includes determining a first vector representation of the posting user and a second vector representation of the recipient user based on one or more features associated with the post, the posting user, and the recipient user. The method still further includes comparing the vector representations and building a machine learning model for automatically detecting spam posts in the social-networking system using a subset of the plurality of posts as non-spam training data.

    Transliteration of text entry across scripts

    公开(公告)号:US11227110B1

    公开(公告)日:2022-01-18

    申请号:US16832089

    申请日:2020-03-27

    Applicant: Facebook, Inc.

    Abstract: Embodiments are disclosed for transliterating text entries across different script systems. A method according to some embodiments includes steps of: receiving an input string in a first script system input using a keyboard; segmenting, using a probabilistic model, the input string into phonemes that correspond to characters or sets of characters in a second script system; converting the phonemes in the first script system into the characters or sets of characters in the second script system, the characters or sets of characters forming a word or a word prefix in the second script system; and outputting the word or the word prefix in the second script system.

    Customizing content based on predicted user preferences

    公开(公告)号:US11226831B2

    公开(公告)日:2022-01-18

    申请号:US15369547

    申请日:2016-12-05

    Applicant: Facebook, Inc.

    Abstract: A system provides content describing an object, for example, a physical object or a customizable service. The system presents an interactive user interface displaying a customized version of a physical object. The interactive user interface allows the user to further customize the physical object. The user interface displays a plurality of options for customizing various components of the physical object. The system customizes the physical object presented via the interactive user interface based on the user profile and user interactions of the user with other users via an online system. The system allows the user to further customize the physical object via the interactive user interface. The interactive user interface presents various options for customization of the physical object based on predicted user preferences.

    IDENTIFYING COOCCURING PROUDUCTS IN CONTENT PROVIDED TO AN ONLINE SYSTEM BY USERS

    公开(公告)号:US20210406980A1

    公开(公告)日:2021-12-30

    申请号:US16917872

    申请日:2020-06-30

    Applicant: Facebook, Inc.

    Abstract: A publishing user identifies a product offered by the user to an online system by providing multiple images of various products viewed at different angles to the online system. The online system applies an identification model to content items obtained from other users to identify one or more of the products in various content items. Based on a product selected by a viewing user, the online system applies the identification model to various content items to identify content items including the selected product and to identify other products included in content items including the selected product. The online system provides information identifying the other products included in content items along with the selected product to the viewing user. The online system may rank other products included in content items along with the selected product and identify other products to the viewing user based on the ranking.

    SYSTEM AND METHODS FOR FEATURE ENGINEERING BASED ON GRAPH LEARNING

    公开(公告)号:US20210406779A1

    公开(公告)日:2021-12-30

    申请号:US16913054

    申请日:2020-06-26

    Applicant: Facebook, Inc.

    Abstract: In one embodiment, a computing system may receive query information associated with a machine-learning model. The system may access a knowledge graph that defines relationships between a number of machine-learning models and a number of features of the machine-learning models. The system may determine, based on the knowledge graph and the query information, one or more correlation metrics indicating correlations between the machine-learning model and one or more features of the features in the knowledge graph. The system may determine one or more recommended features for the machine-learning model based on the one or more correlation metrics and the one or more features.

    Dynamically modifying live video streams for participant devices in digital video rooms

    公开(公告)号:US11202035B2

    公开(公告)日:2021-12-14

    申请号:US17196704

    申请日:2021-03-09

    Applicant: Facebook, Inc.

    Abstract: This disclosure describes a video room system that can efficiently and flexibly generate and facilitate video rooms for video chatting between participant devices by prioritizing video streams to provide to participant devices based on performance parameters. For example, the video room system can provide different video room streams to different participant devices based on the performance parameters of the participant devices. In addition, the video room system can customize or tailor video room streams based on identifying dominant speaker devices, determining connections between participant devices as indicated by a social graph of a social networking system, and/or based on join methods utilized by participant devices to join a video room.

Patent Agency Ranking