知搜-全球专利检索

  1. Login
  2. Sign Up

Search Results

194 records (took 0.021 seconds)

    Efficient method for indexing data transferred between machines in distributed graph processing systems
    22.
    发明授权
    Efficient method for indexing data transferred between machines in distributed graph processing systems 有权

    公开(公告)号：US10002205B2

    公开(公告)日：2018-06-19

    申请号：US14947382

    申请日：2015-11-20

    Applicant: Oracle International Corporation

    Inventor： Nicholas Roth ,  Sungpack Hong ,  Siegfried Depner ,  Thomas Manhardt ,  Hassan Chafi

    IPC: G06F17/30

    CPC classification number: G06F16/9024 ,  G06F16/278

    Abstract: Techniques herein index data transferred during distributed graph processing. In an embodiment, a system of computers divides a directed graph into partitions. The system creates one partition per computer and distributes each partition to a computer. Each computer builds four edge lists that enumerate edges that connect the partition of the computer with a partition of a neighbor computer. Each of the four edge lists has edges of a direction, which may be inbound or outbound from the partition. Edge lists are sorted by identifier of the vertex that terminates or originates each edge. Each iteration of distributed graph analysis involves each computer processing its partition and exchanging edge data or vertex data with neighbor computers. Each computer uses an edge list to build a compactly described range of edges that connect to another partition. The computers exchange described ranges with their neighbors during each iteration.

    Automated generation of memory consumption aware code
    23.
    发明授权
    Automated generation of memory consumption aware code 有权

    公开(公告)号：US09971570B2

    公开(公告)日：2018-05-15

    申请号：US14969231

    申请日：2015-12-15

    Applicant: Oracle International Corporation

    Inventor： Martin Sevenich ,  Sungpack Hong ,  Hassan Chafi

    IPC: G06F9/44

    CPC classification number: G06F8/456

    Abstract: Techniques generate memory-optimization logic for concurrent graph analysis. A computer analyzes domain-specific language logic that analyzes a graph having vertices and edges. The computer detects parallel execution regions that create thread locals. Each thread local is associated with a vertex or edge. For each parallel region, the computer calculates how much memory is needed to store one instance of each thread local. The computer generates instrumentation that determines how many threads are available and how many vertices and edges will create thread locals. The computer generates tuning logic that determines how much memory is originally needed for the parallel region based on how much memory is needed to store the one instance, how many threads are available, and graph size. The tuning logic detects a memory shortage based on the original amount of memory needed exceeding how much memory is available and accordingly adjusts the execution of the parallel region.

    EFFICIENT METHOD FOR SUBGRAPH PATTERN MATCHING
    26.
    发明申请
    EFFICIENT METHOD FOR SUBGRAPH PATTERN MATCHING 审中-公开

    公开(公告)号：US20170169133A1

    公开(公告)日：2017-06-15

    申请号：US14969789

    申请日：2015-12-15

    Applicant: Oracle International Corporation

    Inventor： Jinha Kim ,  Oskar Van Rest ,  Sungpack Hong ,  Hassan Chafi

    IPC: G06F17/30

    CPC classification number: G06F17/30958 ,  G06F17/30324

    Abstract: Techniques herein optimize subgraph pattern matching. A computer receives a graph vertex array and a graph edge array. Each vertex and each edge has labels. The computer stores an array of index entries and an array of edge label sets. Each index entry corresponds to a respective vertex originating an edge and associates an offset of the edge with an offset of the respective vertex. Each edge label set contains labels of a respective edge. The computer selects a candidate subset of edges originating at a current vertex. The edge labels of each candidate edge of the candidate subset include a same particular query edge labels. The computer selects the candidate subset based on the index array and afterwards selects a result subset of vertices from among the terminating vertices of the candidate edges. The labels of each vertex of the result subset include a same particular query vertex labels.

    Automatic Selection Of An Abstract Data Type
    27.
    发明申请
    Automatic Selection Of An Abstract Data Type 审中-公开
    自动选择抽象数据类型

    公开(公告)号：US20150331683A1

    公开(公告)日：2015-11-19

    申请号：US14276895

    申请日：2014-05-13

    Applicant: Oracle International Corporation

    Inventor： Martin Sevenich ,  Sungpack Hong ,  Hassan Chafi

    IPC: G06F9/45

    CPC classification number: G06F8/443 ,  G06F8/437

    Abstract: An implementation of an abstract data type is automatically selected by a compiler. The compiler chooses an implementation for each instance of an abstract data type in a program based on operations performed in the instance within the program.

    Abstract translation: 抽象数据类型的实现由编译器自动选择。 编译器根据程序中的实例执行的操作，在程序中为抽象数据类型的每个实例选择一个实现。

    System For Applying Transformation To Improve Graph Analysis
    28.
    发明申请
    System For Applying Transformation To Improve Graph Analysis 有权
    应用变换的系统来改善图形分析

    公开(公告)号：US20140189665A1

    公开(公告)日：2014-07-03

    申请号：US13733424

    申请日：2013-01-03

    Applicant: ORACLE INTERNATIONAL CORPORATION

    Inventor： Sungpack Hong ,  Hassan Chafi ,  Eric Sedlar

    IPC: G06F9/45

    CPC classification number: G06F8/443

    Abstract: A method for performing a neighbor-flipping transformation is provided. In one embodiment, a graph analysis program for computing a function relating to nodes in a directed graph is obtained and analyzed for neighborhood iterating operations, in which a function is computed over sets of nodes in the graph. For any detected neighborhood iterating operation, the method transforms the iterating operation by reversing the neighbor node relationship between the nodes in the operation. The transformed operation computes the same value for the function as the operation prior to transformation. The method alters the neighbor node relationship automatically, so that a user does not have to recode the graph analysis program. In some cases, the method includes construction of edges in the reverse direction while retaining the original edges in addition to performing the transformation.

    Abstract translation: 提供了一种执行相邻翻转变换的方法。 在一个实施例中，获得用于计算与有向图中的节点有关的功能的图分析程序，并对其进行邻域迭代操作进行分析，其中在图中的节点集合上计算函数。 对于任何检测到的邻域迭代操作，该方法通过在操作中颠倒节点之间的邻居节点关系来转换迭代操作。 转换的操作计算与变换之前的操作相同的函数值。 该方法自动更改邻居节点关系，使用户不必重新编码图形分析程序。 在某些情况下，除了执行变换之外，该方法还包括沿相反方向构造边缘，同时保留原始边缘。

    TRANSFORMER-BASED HYBRID RECOMMENDATION MODEL WITH CONTEXTUAL FEATURE SUPPORT
    29.
    发明申请
    TRANSFORMER-BASED HYBRID RECOMMENDATION MODEL WITH CONTEXTUAL FEATURE SUPPORT 有权

    公开(公告)号：US20250156637A1

    公开(公告)日：2025-05-15

    申请号：US18505293

    申请日：2023-11-09

    Applicant: Oracle International Corporation

    Inventor： Cesare Bernardis ,  Damien Hilloulin ,  Rhicheek Patra ,  Sungpack Hong ,  Hassan Chafi

    IPC: G06F40/284 ,  G06F16/9535

    Abstract: In a computer-implemented embodiment, an interaction machine learning model is trained based on many interactions on many resources. A context lexical token is inferred that represents a current operational context of a user. The context lexical token is inserted into a sequence of other inferred lexical tokens. From the context lexical token within the sequence of tokens, the interaction machine learning model infers a predicted resource that will be accessed next. In an embodiment, accelerated matchmaking entails suitability measurement by a dot product of a) a dynamically inferred user embedding that is based on the context lexical token and b) a statically inferred item embedding.

Patent Agency Ranking