中科微点-全球专利检索

  1. Login
  2. Sign Up

Search Results

25 records (took 0.015 seconds)

    NODE GRAPH PRUNING AND FRESH CONTENT
    2.
    发明公开
    NODE GRAPH PRUNING AND FRESH CONTENT 审中-公开

    公开(公告)号：US20230385338A1

    公开(公告)日：2023-11-30

    申请号：US18364998

    申请日：2023-08-03

    Applicant: Pinterest, Inc.

    Inventor： Chantat Eksombatchai ,  Jurij Leskovec ,  Rahul Sharma ,  Charles Walsh Sugnet ,  Mark Bormann Ulrich

    IPC: G06F16/901 ,  G06F16/435 ,  G06Q30/0201 ,  G06F16/2457

    CPC classification number: G06F16/9024 ,  G06F16/435 ,  G06Q30/0201 ,  G06F16/24578 ,  G06F16/487

    Abstract: This disclosure describes systems and methods that facilitate the generation of recommendations by traversing a graph. Walks that traverse the graph may be initiated from a plurality of different nodes in the node graph. In order to give greater or lesser weight to particular nodes, the walks may have different lengths depending on the nodes from which they are initiated, or an unequal amount of walks may be distributed between nodes from which walks are initiated. A plurality of walks through a node graph may be tracked, and visit counts or scores for nodes in the node graph may be determined. For example, scores may be increased for nodes that are visited by a walk initiated from a first node and a second walk initiated from a second node, or scores may be decreased for nodes that are not visited by a first walk initiated from a first node and a second walk initiated from a second node. Content corresponding to nodes may be recommended based on the scores or visit counts.

    GENERATING NEIGHBORHOOD CONVOLUTIONS WITHIN A LARGE NETWORK
    5.
    发明申请
    GENERATING NEIGHBORHOOD CONVOLUTIONS WITHIN A LARGE NETWORK 审中-公开

    公开(公告)号：US20190286658A1

    公开(公告)日：2019-09-19

    申请号：US16273939

    申请日：2019-02-12

    Applicant: Pinterest, Inc.

    Inventor： Jurij Leskovec ,  Chantat Eksombatchai ,  Kaifeng Chen ,  Ruining He ,  Rex Ying

    IPC: G06F16/901 ,  G06F16/906 ,  G06K9/62 ,  G06F16/22 ,  G06F16/9536 ,  G06N20/00

    Abstract: Systems and methods for generating embeddings for nodes of a corpus graph are presented. More particularly, operations for generation of an aggregated embedding vector for a target node is efficiently divided among operations on a central processing unit and operations on a graphic processing unit. With regard to a target node within a corpus graph, processing by one or more central processing units (CPUs) is conducted to identify the target node's relevant neighborhood (of nodes) within the corpus graph. This information is prepared and passed to one or more graphic processing units (GPUs) that determines the aggregated embedding vector for the target node according to data of the relevant neighborhood of the target node.

    Data reduction for node graph creation
    6.
    发明授权
    Data reduction for node graph creation 有权

    公开(公告)号：US11256747B1

    公开(公告)日：2022-02-22

    申请号：US15870776

    申请日：2018-01-12

    Applicant: Pinterest, Inc.

    Inventor： Chantat Eksombatchai ,  Jurij Leskovec ,  Zitao Liu ,  Rahul Sharma ,  Mark Bormann Ulrich

    IPC: G06F16/00 ,  G06F16/901 ,  G06K9/62 ,  G06F16/958

    Abstract: This disclosure describes systems and methods that facilitate reducing a data set that may be used to construct a node graph. For example, the data set may include collections, representations, and associations between the collections and the representations. Topic scores may be determined for the representations, and diversity scores for each collection may be determined based on the topic scores of representations that are associated with the respective collection. If the diversity score is too high, then the collection and its associations are excluded from being incorporated into a node graph that is subsequently constructed from the data set. Topic scores may also be determined for collections in the data set based on the topic scores of representations that are associated with each collection.

    Node graph traversal methods
    7.
    发明授权
    Node graph traversal methods 有权

    公开(公告)号：US10762134B1

    公开(公告)日：2020-09-01

    申请号：US15870785

    申请日：2018-01-12

    Applicant: Pinterest, Inc.

    Inventor： Chantat Eksombatchai ,  Jurij Leskovec ,  Rahul Sharma ,  Charles Walsh Sugnet ,  Mark Bormann Ulrich

    IPC: G06F7/00 ,  G06F16/901 ,  G06Q30/02 ,  G06F16/2457 ,  G06F16/435 ,  G06F16/487

    Abstract: This disclosure describes systems and methods that facilitate the generation of recommendations by traversing a graph. Walks that traverse the graph may be initiated from a plurality of different nodes in the node graph. In order to give greater or lesser weight to particular nodes, the walks may have different lengths depending on the nodes from which they are initiated, or an unequal amount of walks may be distributed between nodes from which walks are initiated. A plurality of walks through a node graph may be tracked, and visit counts or scores for nodes in the node graph may be determined. For example, scores may be increased for nodes that are visited by a walk initiated from a first node and a second walk initiated from a second node, or scores may be decreased for nodes that are not visited by a first walk initiated from a first node and a second walk initiated from a second node. Content corresponding to nodes may be recommended based on the scores or visit counts.

    Node graph traversal methods
    8.
    发明授权
    Node graph traversal methods 有权

    公开(公告)号：US10740399B1

    公开(公告)日：2020-08-11

    申请号：US15870781

    申请日：2018-01-12

    Applicant: Pinterest, Inc.

    Inventor： Chantat Eksombatchai ,  Jurij Leskovec ,  Pranav Jindal ,  Rahul Sharma ,  Mark Bormann Ulrich

    IPC: G06F17/00 ,  G06F16/901 ,  G06Q30/02 ,  G06F16/435 ,  G06F16/487 ,  G06F16/2457

    Abstract: This disclosure describes systems and methods that facilitate generating recommendations by traversing a node graph. For example, recommendations may be generated for a node in the node graph by running a plurality of walks through the node graph and tracking the nodes visited by the walks. For example, a visit count or score may be maintained and/or updated for each node as the walks traverse through the node graph. The walks may be terminated after a defined amount of nodes in the node graph have visit counts or scores that satisfy a criterion. Content corresponding to nodes with the highest visit counts or scores may be recommended.

    Node graph traversal methods
    9.
    发明授权
    Node graph traversal methods 有权

    公开(公告)号：US10671672B1

    公开(公告)日：2020-06-02

    申请号：US15870790

    申请日：2018-01-12

    Applicant: Pinterest, Inc.

    Inventor： Chantat Eksombatchai ,  Jurij Leskovec

    IPC: G06F16/901 ,  G06Q30/02 ,  G06F16/435 ,  G06F16/2457 ,  G06F16/487

    Abstract: This disclosure describes systems and methods that facilitate generating recommendations by traversing a node graph. For example, a cluster of nodes in a node graph may be determined for a target node in the node graph based at least in part on a proximity of the nodes in the cluster to the target node in the node graph. A plurality of walks through a node graph may be tracked, and a visit count or score for the target node may be increased for each visit to a node in the cluster. The walks may be terminated after a defined amount of walks have been performed or a defined amount of nodes in the node graph have scores that satisfy a criterion. Content corresponding to nodes may be recommended based on scores or visit counts.

    EFFICIENT PROCESSING OF NEIGHBORHOOD DATA
    10.
    发明申请
    EFFICIENT PROCESSING OF NEIGHBORHOOD DATA 审中-公开

    公开(公告)号：US20190286754A1

    公开(公告)日：2019-09-19

    申请号：US16178443

    申请日：2018-11-01

    Applicant: Pinterest, Inc.

    Inventor： Jurij Leskovec ,  Chantat Eksombatchai ,  Kaifeng Chen ,  Ruining He ,  Rex Ying

    IPC: G06F17/30 ,  G06N3/04 ,  G06N3/08

    Abstract: Systems and methods for generating embeddings for nodes of a corpus graph are presented. More particularly, operations for generation of an aggregated embedding vector for a target node is efficiently divided among operations on a central processing unit and operations on a graphic processing unit. With regard to a target node within a corpus graph, processing by one or more central processing units (CPUs) is conducted to identify the target node's relevant neighborhood (of nodes) within the corpus graph. This information is prepared and passed to one or more graphic processing units (GPUs) that determines the aggregated embedding vector for the target node according to data of the relevant neighborhood of the target node.

Patent Agency Ranking