EP-4736025-A1 - GENERATIVE SUMMARIZATION DIALOG-BASED INFORMATION RETRIEVAL SYSTEM

Abstract

Embodiments of the disclosed technologies include generating a search prompt based on an input portion of an online dialog involving a user of a computing device. The search prompt includes a dialog summarization instruction configured to instruct a generative artificial intelligence model to generate and output a dialog summary. The search prompt is sent to a first generative model. In response to the search prompt, a search query is generated and output by the first generative model based on the dialog summary. The search query is sent to a search system. Search result data is determined based on an execution of the search query by the search system. At least some of the search result data is included in an output portion of the online dialog. The output portion is configured to be displayed at the computing device in response to the input portion of the online dialog.

Inventors

• KRISHNAN, APARNA
• LLOYD, CHRISTOPHER WRIGHT, II
• OWEN, Jeremy K.
• FONG, CHRISTOPHER J.
• SUNDARESH, SUMAN
• SHAH, Lavish
• KHURRAM, Muhammad Basit
• JILLINGS, Michaels

Assignees

• Microsoft Technology Licensing, LLC

Dates

Publication Date

20260506

Application Date

20240621

Claims (20)

1. 1. A computer-implemented method comprising: generating (702) a first search prompt based on a first input portion of an online dialog involving a user of a computing device, wherein the first search prompt comprises a dialog summarization instruction to generate and output a dialog summary, and the dialog summary comprises a machine-generated summary based on at least one of a dialog history, attribute data associated with the user, or online activity' data associated with the user; sending (704) the first search prompt to a first large language model; receiving (706) a first search query', wherein, in response to the first search prompt, the first search query is generated and output by the first large language model based on the dialog summary ; sending (708) the first search query to a search system; receiving search result data, wherein the search result data is determined based on an execution of the first search query by' the search system; and including (712) at least some of the search result data in a first output portion of the online dialog, wherein the first output portion is configured to be displayed at the computing device in response to the first input portion of the online dialog.
2. 2. The computer-implemented method of claim 1, wherein generating the first search prompt further comprises: extracting a topic from the dialog summary'; and including, in the first search prompt, a topic matching instruction configured to instruct the first large language model to filter the search result data based on the extracted topic.
3. 3. The computer-implemented method of claim 1, wherein generating the first search prompt further comprises: traversing an entity graph to identify attribute data associated with the user; retrieving at least some of the identified attribute data from at least one data store; and including the retrieved attribute data in the first search prompt, wherein the first search prompt further comprises a query' disambiguation instruction configured to instruct the first large language model to use the retrieved attribute data to disambiguate an ambiguous portion of the dialog history.
4. 4. The computer-implemented method of claim 1, wherein generating the first search prompt further comprises: retrieving stored attribute data associated with the user; and including the retrieved attribute data in the first search prompt, wherein the first search prompt further comprises a query expansion instruction configured to instruct the first large language model to use the retrieved attribute data to expand the first search query.
5. 5. The computer-implemented method of claim 1, wherein generating the first search prompt further comprises: traversing an entity graph to identify online activity data associated with the user; retrieving at least some of the identified online activity data from at least one data store; and including the retrieved online activity data in the first search prompt, wherein the first search prompt further comprises a query disambiguation instruction configured to instruct the first large language model to use the retrieved online activity data to disambiguate an ambiguous portion of the dialog history.
6. 6. The computer-implemented method of claim 1, wherein generating the first search prompt further comprises: retrieving stored online activity data associated with the user; and including the retrieved online activity data in the first search prompt, wherein the first search prompt further comprises a query expansion instruction configured to instruct the first large language model to use the retrieved online activity data to expand the first search query.
7. 7. The computer-implemented method of claim 1, further comprising: generating a first response prompt based on the first input portion of the online dialog, the dialog summary, and the search result data; sending the first response prompt to a second large language model; receiving a first response, wherein the first response is generated and output by the second large language model based on the first response prompt; and including the first response in the first output portion of the online dialog.
8. 8. The computer-implemented method of claim 7, wherein generating the first response prompt further comprises: including, in the first response prompt, a result summarization instruction configured to instruct the second large language model to generate and output a result summary of the search result data, wherein the first response is based on the result summary.
9. 9. The computer-implemented method of claim 7, wherein generating the first response prompt further comprises: including, in the first response prompt, a relevance instruction configured to instruct the second large language model to determine relevance of the search result data to the first input portion of the online dialog and include relevant search result data in the first response.
10. 10. The computer-implemented method of claim 1, wherein receiving the search result data comprises at least one of: traversing an entity graph to determine at least one recommendation that matches the first search query, wherein the at least one recommendation comprises at least one of an online resource or a human resource; or traversing an index to identify at least one digital content item that matches the first search query', wherein the at least one digital content item comprises at least one of an article, a document, an audio file, or a video file; or receiving, from at least one machine learning model, at least one recommendation that matches the first search query.
11. 11. A system comprising: at least one processor; and at least one memory device coupled to the at least one processor, wherein the at least one memory device comprises at least one instruction that, when executed by the at least one processor, cause the at least one processor to perform at least one operation comprising: generating (702) a first search prompt based on a first input portion of an online dialog involving a user of a computing device, wherein the first search prompt comprises a dialog summarization instruction to generate and output a dialog summary, and the dialog summary comprises a machine-generated summary based on at least one of a dialog history, attribute data associated with the user, or online activity' data associated with the user; sending (704) the first search prompt to a first large language model; receiving (706) a first search query', wherein, in response to the first search prompt, the first search query is generated and output by the first large language model based on the dialog summary'; sending (708) the first search query to a search system; receiving (710) search result data, wherein the search result data is determined based on an execution of the first search query by the search system; and including (712) at least some of the search result data in a first output portion of the online dialog, wherein the first output portion is configured to be displayed at the computing device in response to the first input portion of the online dialog.
12. 12. The system of claim 11, wherein generating the first search prompt further comprises at least one of: (a) (i) extracting a topic from the dialog summary'; and (ii) including, in the first search prompt, a topic matching instruction configured to instruct the first large language model to filter the search result data based on the extracted topic; or (b) (i) traversing an entity graph to identify attribute data associated with the user; (ii) retrieving at least some of the identified attribute data from at least one data store; and (iii) including the retrieved attribute data in the first search prompt, wherein the first search prompt further comprises a query 7 disambiguation instruction configured to instruct the first large language model to use the retrieved attribute data to disambiguate an ambiguous portion of the dialog history; or (c) (i) retrieving stored attribute data associated with the user; and (ii) including the retrieved attribute data in the first search prompt, wherein the first search prompt further comprises a query expansion instruction configured to instruct the first large language model to use the retrieved attribute data to expand the first search query; or (d) (i) traversing an entity graph to identify online activity' data associated with the user; (ii) retrieving at least some of the identified online activity' data from at least one data store; and (iii) including the retrieved online activity data in the first search prompt, wherein the first search prompt further comprises a query disambiguation instruction configured to instruct the first large language model to use the retrieved online activity data to disambiguate an ambiguous portion of the dialog history; or (e) (i) retrieving stored online activity data associated with the user; and (ii) including the retrieved online activity data in the first search prompt, wherein the first search prompt further comprises a query expansion instruction configured to instruct the first large language model to use the retrieved online activity' data to expand the first search query.
13. 13. The system of claim 11, wherein the instructions, when executed by the at least one processor, cause the at least one processor to perform at least one operation further comprising: generating a first response prompt based on the first input portion of the online dialog, the dialog summary, and the search result data; sending the first response prompt to a second large language model; receiving a first response, wherein the first response is generated and output by' the second large language model based on the first response prompt; and including the first response in the first output portion of the online dialog.
14. 14. The system of claim 13, wherein generating the first response prompt further comprises at least one of: including, in the first response prompt, a result summarization instruction configured to instruct the second large language model to generate and output a result summary of the search result data, wherein the first response is based on the result summary; or including, in the first response prompt, a relevance instruction configured to instruct the second large language model to determine relevance of the search result data to the first input portion of the online dialog and include relevant search result data in the first response.
15. 15. The system of claim 11, wherein receiving the search result data comprises at least one of: traversing an entity graph to determine at least one recommendation that matches the first search query, wherein the at least one recommendation comprises at least one of an online resource or a human resource; or traversing an index to identify at least one digital content item that matches the first search query, wherein the at least one digital content item comprises at least one of an article, a document, an audio file, or a video file; or receiving, from at least one machine learning model, at least one recommendation that matches the first search query.
16. 16. At least one non-transitory machine readable storage medium comprising at least one instruction that, when executed by at least one processor, cause the at least one processor to perform at least one operation comprising: generating (702) a first search prompt based on a first input portion of an online dialog involving a user of a computing device, wherein the first search prompt comprises a dialog summarization instruction to generate and output a dialog summary, and the dialog summary comprises a machine-generated summary based on at least one of a dialog history, attribute data associated with the user, or online activity' data associated with the user; sending (704) the first search prompt to a first large language model; receiving (706) a first search query', wherein, in response to the first search prompt, the first search query' is generated and output by the first large language model based on the dialog summary ; sending (708) the first search query to a search system; receiving (710) search result data, wherein the search result data is determined based on an execution of the first search query by the search system; and including (712) at least some of the search result data in a first output portion of the online dialog, wherein the first output portion is configured to be displayed at the computing device in response to the first input portion of the online dialog.
17. 17. The at least one non-transitory machine readable storage medium of claim 16, wherein generating the first search prompt further comprises at least one of: (a) (i) extracting a topic from the dialog summary; and (ii) including, in the first search prompt, a topic matching instruction configured to instruct the first large language model to filter the search result data based on the extracted topic; or (b) (i) traversing an entity graph to identify attribute data associated with the user; (ii) retrieving at least some of the identified attribute data from at least one data store; and (iii) including the retrieved attribute data in the first search prompt, wherein the first search prompt further comprises a query disambiguation instruction configured to instruct the first large language model to use the retrieved attribute data to disambiguate an ambiguous portion of the dialog history; or (c) (i) retrieving stored attribute data associated with the user; and (ii) including the retrieved attribute data in the first search prompt, wherein the first search prompt further comprises a query expansion instruction configured to instruct the first large language model to use the retrieved attribute data to expand the first search query; or (d) (i) traversing an entity graph to identify online activity data associated with the user; (ii) retrieving at least some of the identified online activity data from at least one data store; and (iii) including the retrieved online activity data in the first search prompt, wherein the first search prompt further comprises a query disambiguation instruction configured to instruct the first large language model to use the retrieved online activity data to disambiguate an ambiguous portion of the dialog history; or (e) (i) retrieving stored online activity data associated with the user; and (ii) including the retrieved online activity data in the first search prompt, wherein the first search prompt further comprises a query expansion instruction configured to instruct the first large language model to use the retrieved online activity data to expand the first search query.
18. 18. The at least one non-transitory machine readable storage medium of claim 16, wherein the instructions, when executed by the at least one processor, cause the at least one processor to perform at least one operation further comprising: generating a first response prompt based on the first input portion of the online dialog, the dialog summary, and the search result data; sending the first response prompt to a second large language model; receiving a first response, wherein the first response is generated and output by the second large language model based on the first response prompt; and including the first response in the first output portion of the online dialog.
19. 19. The at least one non-transitory machine readable storage medium of claim 18, wherein generating the first response prompt further comprises at least one of: including, in the first response prompt, a result summarization instruction configured to instruct the second large language model to generate and output a result summary of the search result data, wherein the first response is based on the result summary; or including, in the first response prompt, a relevance instruction configured to instruct the second large language model to determine relevance of the search result data to the first input portion of the online dialog and include relevant search result data in the first response.
20. 20. The at least one non-transitory machine readable storage medium of claim 16, wherein receiving the search result data comprises at least one of: traversing an entity graph to determine at least one recommendation that matches the first search query, wherein the at least one recommendation comprises at least one of an online resource or a human resource; or traversing an index to identify at least one digital content item that matches the first search query, wherein the at least one digital content item comprises at least one of an article, a document, an audio file, or a video file; or receiving, from at least one machine learning model, at least one recommendation that matches the first search query.

Description

GENERATIVE SUMMARIZATION DIALOG-BASED INFORMATION RETRIEVAL SYSTEM TECHNICAL FIELD [001] A technical field to which the present disclosure relates includes computer programs that use artificial intelligence to understand user queries and automate responses to those queries in a manner that simulates human conversation. Another technical field to which the present disclosure relates is generative artificial intelligence. COPYRIGHT NOTICE [002] This patent document, including the accompanying drawings, contains material that is subject to copyright protection. The copyright owner has no objection to the facsimile reproduction of this patent document, as it appears in the publicly accessible records of the United States Patent and Trademark Office, consistent with the fair use principles of the United States copyright laws, but otherwise reserves all copyright rights whatsoever. BACKGROUND [003] A search engine is a software system that is designed to find and retrieve stored information that matches a search query. A chatbot (or chat bot) is a software application that can retrieve information and answer questions by simulating a natural language conversation with a human user. A recommendation system is a software system that automatically generates proactive recommendations for a user without explicitly receiving a query' from the user. BRIEF DESCRIPTION OF THE DRAWINGS [004] The disclosure will be understood more fully from the detailed description given below and from the accompanying drawings of various embodiments of the disclosure. The drawings are for explanation and understanding only and should not be taken to limit the disclosure to the specific embodiments shown. [005] FIG. 1A is a flow diagram of an example method for generative summarization dialogbased information retrieval using components of a generative summarization dialog-based information retrieval system in accordance with some embodiments of the present disclosure. [006] FIG. IB is a block diagram of an example of components of a search prompt for a generative summarization dialog-based information retrieval system in accordance with some embodiments of the present disclosure. [007] FIG. 1C is a block diagram of an example of components of a search system that can be used in connection with a generative summarization dialog-based information retrieval system in accordance with some embodiments of the present disclosure. [008] FIG. ID is a block diagram of an example of components of a response prompt that can be used in connection with a generative summarization dialog-based information retrieval system in accordance with some embodiments of the present disclosure. [009] FIG. 2A is a timing diagram showing an example of communications between dialogbased information retrieval interface and components of a generative summarization dialogbased information retrieval system in accordance with some embodiments of the present disclosure. [0010] FIG. 2B is a flow diagram showing an example of communications among a search system and a generative summarization dialog-based information retrieval system in accordance with some embodiments of the present disclosure. [0011] FIG. 2C is a flow diagram showing an example of communications among a search system and a generative summarization dialog-based information retrieval system in accordance with some embodiments of the present disclosure. [0012] FIG. 3A, FIG. 3B, FIG. 3C. FIG. 3D, and FIG. 3E illustrate an example of at least one flow including screen captures of user interface screens configured to provide generative summarization dialog-based information retrieval in accordance with some embodiments of the present disclosure. [0013] FIG. 4A, FIG. 4B, and FIG. 4C illustrate an example of at least one flow including screen captures of user interface screens configured to provide generative summarization dialogbased information retrieval in accordance with some embodiments of the present disclosure. [0014] FIG. 5 is a block diagram of a computing system that includes a generative summarization dialog-based information retrieval system in accordance with some embodiments of the present disclosure. [0015] FIG. 6 is an example of an entity graph in accordance with some embodiments of the present disclosure. [0016] FIG. 7 is a flow diagram of an example method for generative summarization dialogbased information retrieval using components of a generative summarization dialog-based information retrieval system in accordance with some embodiments of the present disclosure. [0017] FIG. 8 is a block diagram of an example computer system including components of a generative summarization dialog-based information retrieval system in accordance with some embodiments of the present disclosure. DETAILED DESCRIPTION [0018] People often turn to their computing devices when they need an answer to a question or would like to learn about a particular topic. For example, new issues frequently arise in the workplac