公开(公告)号：US20180285170A1

公开(公告)日：2018-10-04

申请号：US15582242

申请日：2017-04-28

Applicant: Microsoft Technology Licensing, LLC

Inventor： Michael Gamon ,  Mark Encarnacion ,  Patrick Pantel ,  Ahmed Hassan Awadallah ,  Madian Khabsa ,  Yu Su

IPC: G06F9/54 ,  G06F17/27 ,  G06N7/00

Abstract: Subject matter involves using natural language to Web application program interfaces (API), which map natural language commands into API calls, or API commands. This mapping enables an average user with little or no programming expertise to access Web services that use API calls using natural language. An API schema is accessed and using a specialized grammar, with the help of application programmers, canonical commands associated with the API calls are generated. A hierarchical probabilistic distribution may be applied to a semantic mesh associated with the canonical commands to identify elements of the commands that require labeling. The identified elements may be sent to annotators, for labeling with NL phrases. Labeled elements may be applied to the semantic mesh and probabilities, or weights updated. Labeled elements may be mapped to the canonical commands with machine learning to generate a natural language to API interface. Other embodiments are described and claimed.

公开(公告)号：US10963318B2

公开(公告)日：2021-03-30

申请号：US16654340

申请日：2019-10-16

Applicant: Microsoft Technology Licensing, LLC

Inventor： Ahmed Hassan Awadallah ,  Mark Encarnacion ,  Michael Gamon ,  Madian Khabsa ,  Patrick Pantel ,  Yu Su

IPC: G06F9/54 ,  G06N7/00 ,  G06F40/30

Abstract: Subject matter involves using natural language to Web application program interfaces (API), which map natural language commands into API calls, or API commands. This mapping enables an average user with little or no programming expertise to access Web services that use API calls using natural language. An API schema is accessed and using a specialized grammar, with the help of application programmers, canonical commands associated with the API calls are generated. A hierarchical probabilistic distribution may be applied to a semantic mesh associated with the canonical commands to identify elements of the commands that require labeling. The identified elements may be sent to annotators, for labeling with NL phrases. Labeled elements may be applied to the semantic mesh and probabilities, or weights updated. Labeled elements may be mapped to the canonical commands with machine learning to generate a natural language to API interface. Other embodiments are described and claimed.

公开(公告)号：US20180191862A1

公开(公告)日：2018-07-05

申请号：US15393479

申请日：2016-12-29

Applicant: Microsoft Technology Licensing, LLC

Inventor： Mark Encarnacion ,  Ievgeniia Zhovtobriukh ,  Patrick Pantel ,  Ahmed Awadallah ,  Chetan Bansal ,  Michael Gamon ,  Cem Aykan ,  Michele Banko ,  Mike Snow ,  Johannes Gehrke

IPC: H04L29/08

CPC classification number: H04L67/327 ,  G06Q10/107 ,  H04L67/2804

Abstract: Systems and methods are presented for detecting an action intent within received content, identifying an action completion bot for carrying out the corresponding action, and initiating the action through an action request to the action completion bot. An action delegation agent executing on a computer system, receives notice of received content, where the action delegation agent is not the target of the received content. An analysis of the received content is conducted to identify an action intent of the received content. Based on the action intent, an action registry is consulted to identify a corresponding action completion bot for carrying out the intended action. A request is submitted to the action completion hot to carry out the action.

公开(公告)号：US20200050500A1

公开(公告)日：2020-02-13

申请号：US16654340

申请日：2019-10-16

Applicant: Microsoft Technology Licensing, LLC

Inventor： Ahmed Hassan Awadallah ,  Mark Encarnacion ,  Michael Gamon ,  Madian Khabsa ,  Patrick Pantel ,  Yu Su

IPC: G06F9/54 ,  G06N7/00 ,  G06F17/27

Abstract: Subject matter involves using natural language to Web application program interfaces (API), which map natural language commands into API calls, or API commands. This mapping enables an average user with little or no programming expertise to access Web services that use API calls using natural language. An API schema is accessed and using a specialized grammar, with the help of application programmers, canonical commands associated with the API calls are generated. A hierarchical probabilistic distribution may be applied to a semantic mesh associated with the canonical commands to identify elements of the commands that require labeling. The identified elements may be sent to annotators, for labeling with NL phrases. Labeled elements may be applied to the semantic mesh and probabilities, or weights updated. Labeled elements may be mapped to the canonical commands with machine learning to generate a natural language to API interface. Other embodiments are described and claimed.

公开(公告)号：US20180262457A1

公开(公告)日：2018-09-13

申请号：US15455061

申请日：2017-03-09

Applicant: Microsoft Technology Licensing, LLC

Inventor： Chetan Bansal ,  Mark Encarnacion ,  Patrick Pantel ,  Woon Kiat Wong

IPC: H04L12/58 ,  H04L29/06

Abstract: Electronic messages are automatically debugged. An electronic message sent by a sender to one or more recipients is received and analyzed to identify any issues in the message. Whenever one or more issues are identified in the message, a reply message may be sent to the sender that describes each of the identified issues and includes a proposed fix for each of the identified issues. Each of the issues identified in the message may also be fixed resulting in a fixed version of the message, and the fixed version of the message may be sent to the sender for their approval. The fixed version of the message may also be sent to each of the recipients. Whenever no issues are identified in the message, a reply message may be sent to the sender informing them that no issues were found.