公开(公告)号：US20210224287A1

公开(公告)日：2021-07-22

申请号：US16744834

申请日：2020-01-16

Applicant: Oracle International Corporation

Inventor： Atif Chaudhry ,  Chandrasekharan Iyer ,  Beda Christoph Hammerschmidt

IPC: G06F16/25 ,  G06F16/21 ,  G06F16/22 ,  G06F16/23 ,  G06F16/2457

Abstract: In an artificial neural network, integrality refers to the degree to which a neuron generates, for a given set of inputs, outputs that are near the border of the output range of a neuron. From each neural network of a pool of trained neural networks, a group of neurons with a higher integrality is selected to form a neural network tunnel (“tunnel”). The tunnel must include all input neurons and output neurons from the neural network, and some of the hidden neurons. Tunnels generated from each neural network in a pool are merged to form another neural network. The new network may then be trained.

公开(公告)号：US20250094398A1

公开(公告)日：2025-03-20

申请号：US18885531

申请日：2024-09-13

Applicant: Oracle International Corporation

Inventor： Aleksandra Czarlinska ,  Saurabh Naresh Netravalkar ,  Denis B. Mukhin ,  Harichandan Roy ,  Zhen Hua Liu ,  Sebastian de la Hoz Luna ,  Beda Christoph Hammerschmidt ,  George R. Krupka ,  Bo Xia ,  David Chih-Wei Jiang

IPC: G06F16/22 ,  G06F16/93

Abstract: Techniques for a unified relational database framework for hybrid vector search are provided. In one technique, multiple documents are accessed and a vector table and a text table are generated. For each accessed document, data within the document is converted to plaintext, multiple chunks are generated based on the plaintext, an embedding model generates a vector for each of the chunks, the vectors are stored in the vector table along with a document identifier that identifies the accessed document, tokens are generated based on the plaintext, the tokens are stored in the text table along with the document identifier. Such processing may be performed in a database system in response to a single database statement to create a hybrid index. In response to receiving a hybrid query, a vector query and a text query are generated and executed and the respective results may be combined.

公开(公告)号：US20240126763A1

公开(公告)日：2024-04-18

申请号：US18099034

申请日：2023-01-19

Applicant: Oracle International Corporation

Inventor： Tirthankar Lahiri ,  Juan R. Loaiza ,  Beda Christoph Hammerschmidt ,  Andrew Witkowski ,  Sankar Subramanian ,  Sabina Petride ,  Ajit Mylavarapu ,  Gerald Venzl

IPC: G06F16/2455 ,  G06F16/22

CPC classification number: G06F16/24565 ,  G06F16/221 ,  G06F16/24554

Abstract: Disclosed herein are techniques for storing, within a database system, metadata that indicates an intended usage (IU). Once created, an IU may be assigned to a column to (a) indicate how the column is intended to be used, and (b) affect how the database server behaves when database operations involve values from the column. The IU assigned to a column supplements, but does not replace, the datatype definition for the column. Each IU may have an IU-bundle. The IU-bundle of an IU indicates how the database server behaves with respect to any column that is assigned the IU. For example, the IU-bundle may indicate constraints that the database server must validate during operations on values from columns assigned to the IU. Techniques are also described for implementing multi-column IUs and flexible IUs.

公开(公告)号：US20200210398A1

公开(公告)日：2020-07-02

申请号：US16690817

申请日：2019-11-21

Applicant: Oracle International Corporation

Inventor： Zhen Hua Liu ,  Vikas Arora ,  Ying Lu ,  Sriram Krishnamurthy ,  Hui Joe Chang ,  Beda Christoph Hammerschmidt ,  Joshua Jeffrey Spiegel

IPC: G06F16/22 ,  G06F16/2455 ,  G06F16/28

Abstract: The present invention relates to autonomous tuning of a data grid of documents in a database. Herein are techniques for storage cells to autonomously maintain local indices and other optimization metadata and algorithms to accelerate selective access into a distributed collection of documents. In an embodiment, each storage cell persists a respective subset of documents. Each storage cell stores, in memory, respective index(s) that map each item to location(s), in one or more documents of the respective subset of documents, where the item occurs. One or more computers execute, based on at least a subset of the indices of the storage cells, a data access request from a database management system. In an embodiment, a cloud of JSON document services provides an easy-to-use, fully autonomous JSON document database that horizontally and elastically scales to deliver fast execution of document transactions and queries without needing tuning by a database administrator.

公开(公告)号：US20240119031A1

公开(公告)日：2024-04-11

申请号：US18115629

申请日：2023-02-28

Applicant: Oracle International Corporation

Inventor： Tirthankar Lahiri ,  Srikrishnan Suresh ,  Beda Christoph Hammerschmidt ,  Adrian Daniel Popescu ,  Jesse Kamp ,  Zhen Hua Liu

IPC: G06F16/21 ,  G06F16/23 ,  G06F16/25

CPC classification number: G06F16/212 ,  G06F16/2365 ,  G06F16/258

Abstract: A computer analyzes a relational schema of a database to generate a data entry schema and encodes the data entry schema as JSON. The data entry schema is sent to a database client so that the client can validate entered data before the entered data is sent for storage. From the client, entered data is received that conforms to the data entry schema because the client used the data entry schema to validate the entered data before sending the data. Into the database, the entered data is stored that conforms to the data entry schema. The data entry schema and the relational schema have corresponding constraints on a datum to be stored, such as a range limit for a database column or an express set of distinct valid values. A constraint may specify a format mask or regular expression that values in the column should conform to, or a correlation between values of multiple columns.

公开(公告)号：US11640380B2

公开(公告)日：2023-05-02

申请号：US17198159

申请日：2021-03-10

Applicant: Oracle International Corporation

Inventor： Hui Zhang ,  Zhen Hua Liu ,  Ying Lu ,  Beda Christoph Hammerschmidt ,  Douglas James McMahon ,  Denis B. Mukhin

IPC: G06F16/22

Abstract: Herein is acceleration for JavaScript object notation (JSON) documents in a relational database based on multilevel, multifield, multivalued indexing and querying. In an embodiment, into a table in a relational database, a computer stores many hierarchical data objects that respectively contain multiple levels that respectively contain one or more fields. An index is generated for indexed values in fields in at least two of the multiple levels. Based on the index, a database statement that references the table in the relational database and at least one of the indexed fields is executed. Indices herein provide matching to more fields in more levels of hierarchical data objects thereby decreasing or eliminating further filtering by brute force after index access. Matching ordinal positions within an array field needs no further filtration after index access. A single index provides increased spatial locality of index entries for acceleration. Index maintenance is minimized for acceleration.

公开(公告)号：US20220229971A1

公开(公告)日：2022-07-21

申请号：US17715384

申请日：2022-04-07

Applicant: Oracle International Corporation

Inventor： Zhen Hua Liu ,  Beda Christoph Hammerschmidt ,  Douglas James McMahon ,  Joshua Jeffrey Spiegel

IPC: G06F40/146 ,  G06F16/22 ,  G06F16/23

Abstract: Herein are fine grained updates to pieces of JavaScript object notation (JSON) documents by database statements that can update, delete, and insert parts of JSON documents. In an embodiment, a computer receives a request that specifies a modification of a JSON document that is stored in a compressed format in persistent storage. The modification adds additional content to the JSON document, and/or replaces an old value in the JSON document with a new value that is not a same size as the old value. The modification is recorded in a change log. The change log is eventually applied to the compressed format of the JSON document in the persistent storage without entirely rewriting the compressed format of the JSON document in the persistent storage.

公开(公告)号：US12298946B2

公开(公告)日：2025-05-13

申请号：US17966736

申请日：2022-10-14

Applicant: Oracle International Corporation

Inventor： Zhen Hua Liu ,  Juan R. Loaiza ,  Sundeep Abraham ,  Shubha Bose ,  Hui Joe Chang ,  Shashank Gugnani ,  Beda Christoph Hammerschmidt ,  Tirthankar Lahiri ,  Ying Lu ,  Douglas James McMahon ,  Aurosish Mishra ,  Ajit Mylavarapu ,  Sukhada Pendse ,  Ananth Raghavan

IPC: G06F16/20 ,  G06F16/21 ,  G06F16/2453 ,  G06F16/84

Abstract: JSON Duality Views are object views that return JDV objects. JDV objects are virtual because they are not stored in a database as JSON objects. Rather, JDV objects are stored in shredded form across tables and table attributes (e.g. columns) and returned by a DBMS in response to database commands that request a JDV object from a JSON Duality View. Through JSON Duality Views, changes to the state of a JDV object may be specified at the level of a JDV object. JDV objects are updated in a database using optimistic lock.

公开(公告)号：US11550785B2

公开(公告)日：2023-01-10

申请号：US16389304

申请日：2019-04-19

Applicant: ORACLE INTERNATIONAL CORPORATION

Inventor： Beda Christoph Hammerschmidt ,  Zhen Hua Liu ,  Vikas Arora ,  Chandrasekharan Iyer ,  Beethoven Cheng ,  Ying Hu ,  Douglas James McMahon

IPC: G06F16/242 ,  G06F16/25 ,  G06F16/84

Abstract: Described is a system, method, and computer program product to perform bi-directional mapping of hierarchical data (e.g. JSON, XML) to database object types (e.g., user defined database object types).

公开(公告)号：US11409741B2

公开(公告)日：2022-08-09

申请号：US15993438

申请日：2018-05-30

Applicant: Oracle International Corporation

Inventor： Beda Christoph Hammerschmidt ,  Zhen Hua Liu ,  Vikas Arora ,  Prakashkumar Thiagarajan ,  Douglas James McMahon

IPC: G06F16/00 ,  G06F16/2453 ,  G06F16/22 ,  G06F16/80 ,  G06F16/25

Abstract: Functionalities of new data types are enabled for data stored using native data types (“recognized data types”) supported by DBMSs. Values having a latent data type may be stored using a recognized data type of a DBMS. That is, a value stored in a database table may be recognized by a DBMS as having a recognized data type (e.g., character string) but the value may also include information that indicates the value conforms to a particular latent data type. DBMSs compile database statements. The database statement can contain multiple expressions that may specify a latent data type or that may have a recognized data type but that also imply latent data types. By determining that various expressions have latent data types, operations that rely on the expressions as having a latent data type may be validated in a query and executed accordingly.