UNIFIED SEMANTIC MODEL OF USER INTENTIONS
    11.
    发明公开

    公开(公告)号:US20230418625A1

    公开(公告)日:2023-12-28

    申请号:US18462953

    申请日:2023-09-07

    Applicant: SAP SE

    Inventor: Philip Miseldine

    CPC classification number: G06F9/4451 G06F3/011 G06N20/00 G06F11/3438

    Abstract: The present disclosure involves systems, software, and computer implemented methods for generating a unified semantic model identifying a user's intentions or goals within operating the system. By determining the intent and desires of the user, tasks and queries associated with or requested by the user can be more effectively optimized, automated, or otherwise improved. In some computing environments, multiple users access a database which includes a collection of software objects. The objects within the database can sometimes form a hierarchy, with a parent-child relationship among them. Traditionally, a developer creating the database and the objects within can only determine a user's intent based on the actions taken by the user within the database. The present solution provides the developer with a higher level model representing different users' intents based on observed data from a number of sources in addition to the user's interaction with the database.

    Unified semantic model of user intentions

    公开(公告)号:US11775318B2

    公开(公告)日:2023-10-03

    申请号:US17100472

    申请日:2020-11-20

    Applicant: SAP SE

    Inventor: Philip Miseldine

    CPC classification number: G06F9/4451 G06F3/011 G06F11/3438 G06N20/00

    Abstract: The present disclosure involves systems, software, and computer implemented methods for generating a unified semantic model identifying a user's intentions or goals within operating the system. By determining the intent and desires of the user, tasks and queries associated with or requested by the user can be more effectively optimized, automated, or otherwise improved. In some computing environments, multiple users access a database which includes a collection of software objects. The objects within the database can sometimes form a hierarchy, with a parent-child relationship among them. Traditionally, a developer creating the database and the objects within can only determine a user's intent based on the actions taken by the user within the database. The present solution provides the developer with a higher level model representing different users' intents based on observed data from a number of sources in addition to the user's interaction with the database.

    INTELLIGENT NATURAL LANGUAGE QUERY PROCESSOR
    14.
    发明申请

    公开(公告)号:US20190197185A1

    公开(公告)日:2019-06-27

    申请号:US15852084

    申请日:2017-12-22

    Applicant: SAP SE

    CPC classification number: G06F16/3344 G06F16/35 G06N7/005 G06N20/00

    Abstract: A system for answering a natural language query (NLQ) directed to a database system. The system includes a query receiver receives the NLQ and models the NLQ as a tuple of entities including at least a subject entity and an intention entity. The system further includes a service discovery component and a query transformation unit. The service discovery component discovers a specific database system asset for answering the NLQ by evaluating system model descriptions of database system assets against the NLQ. The query transformation unit transforms the NLQ into a native service call for the specific database system asset, and invokes the native service call for callback of the specific database system asset to answer the NLQ.

Patent Agency Ranking