Retrieval Augmented Clarification in Interactive Systems

    公开(公告)号:US20240202189A1

    公开(公告)日:2024-06-20

    申请号:US18081107

    申请日:2022-12-14

    IPC分类号: G06F16/2452

    CPC分类号: G06F16/24526 G06F16/24522

    摘要: Techniques for generating a clarification to distinguish among retrieved content in interactive systems are provided. In one aspect, a method for generating a clarification prompt in an interactive system includes: obtaining a training dataset for generating the clarification prompt from existing question-answering datasets by modifying original queries in the existing question-answering datasets to obtain training examples of under-specified queries; and training a machine learning model using the training dataset how to select latent differentiating factors in content candidates obtained from an under-specified query from a user and, based on the latent differentiating factors, generate the clarification prompt to clarify an intent of the user.

    Independent object generator and wrapper engine

    公开(公告)号:US11789947B2

    公开(公告)日:2023-10-17

    申请号:US17317013

    申请日:2021-05-11

    摘要: Aspects of the disclosure relate to a data wrapper engine. A computing platform may receive a query comprising a request for data stored as a CLOB. The computing platform may obtain, from a data storage system, the data stored as a CLOB. The computing platform may generate a file wrapper for the data, wherein generating the file wrapper comprises converting the CLOB to a VARCHAR object and storing the VARCHAR object in the file wrapper. The computing platform may generate, using the VARCHAR object stored in the file wrapper, a SQL response to the query. The computing platform may execute the dynamic SQL response to generate a response to the query. The computing platform may send, to a user device, the response to the query and commands directing the user device to display the response to the query, which may cause the user device to display the response.

    Automatic object inference in a database system

    公开(公告)号:US11698899B2

    公开(公告)日:2023-07-11

    申请号:US16952585

    申请日:2020-11-19

    申请人: Dan Benanav

    发明人: Dan Benanav

    摘要: A binary relational database model is described whereby application-layer object structures are easily inferred from database query templates. The object structures take the form of acyclic hypergraphs, which are induced from primal graphs representing query templates. Database applications may iterate through the collection of returned object structures, accessing the data in each structure. The returned object structures are not based on a fixed object model, thereby permitting rich structures with greater applicability than traditional ORM systems. A relationship between non-primitive entities may be directly expressed without the need for alternative join tables. Development and maintenance costs are thus substantially reduced, and data is more efficiently stored and manipulated for database applications.

    ENHANCED DOCUMENT VISUAL QUESTION ANSWERING SYSTEM VIA HIERARCHICAL ATTENTION

    公开(公告)号:US20230153531A1

    公开(公告)日:2023-05-18

    申请号:US17528972

    申请日:2021-11-17

    申请人: ADOBE INC.

    摘要: Systems and methods for performing Document Visual Question Answering tasks are described. A document and query are received. The document encodes document tokens and the query encodes query tokens. The document is segmented into nested document sections, lines, and tokens. A nested structure of tokens is generated based on the segmented document. A feature vector for each token is generated. A graph structure is generated based on the nested structure of tokens. Each graph node corresponds to the query, a document section, a line, or a token. The node connections correspond to the nested structure. Each node is associated with the feature vector for the corresponding object. A graph attention network is employed to generate another embedding for each node. These embeddings are employed to identify a portion of the document that includes a response to the query. An indication of the identified portion of the document is be provided.