公开(公告)号：WO2023279038A1

公开(公告)日：2023-01-05

申请号：PCT/US2022/073279

申请日：2022-06-30

Applicant: PRICEWATERHOUSECOOPERS LLP

Inventor： LI, Chung-Sheng ,  CHENG, Winnie ,  FLAVELL, Mark John ,  HALLMARK, Lori Marie ,  LIZOTTE, Nancy Alayne ,  LEONG, Kevin Ma

IPC: G06F40/10 ,  G06F40/20 ,  G06F40/30 ,  G06N20/00 ,  G10L15/06 ,  G06F40/295 ,  G06K9/62 ,  G06N5/02 ,  G10L15/18 ,  G10L15/197 ,  G06F16/3347 ,  G06F16/353 ,  G06F16/93 ,  G06N5/022 ,  G06N5/041 ,  G06N5/045 ,  G06Q10/0635 ,  G06Q30/018 ,  G06Q40/12 ,  G06V30/412 ,  G06V30/416

Abstract: Systems and methods for adjudicating Al-augmented automated analysis of documents in order to quickly and efficiently make various adjudications based on the documents are provided, including adjudications as to whether the documents represent underlying data that meets one or more predefined or dynamically-determined criteria. Criteria for adjudication may include commercial -sub stance criteria, related-party-transaction criteria, and/or collectability criteria. A system may receive a plurality of documents and generate a plurality of feature vectors by applying natural language processing techniques. The system may apply one or more classification models to the plurality of feature vectors to generate output data classifying each of the feature vectors. The system may identify, for each feature vector, a subset of closest matching prior feature vectors. Based on the classification and based on the identified subset, the system may adjudicate each feature vector with respect to commercial substance, including an adjudication classification and an adjudication confidence score.

公开(公告)号：WO2023275890A1

公开(公告)日：2023-01-05

申请号：PCT/IN2022/050593

申请日：2022-06-28

Applicant: SALINS, Paul Christadas

Inventor： SALINS, Paul Christadas

IPC: G06F16/30 ,  G06F40/10 ,  G06F16/93

Abstract: A method and system for extracting signature patterns from large datasets is disclosed. The method includes encoding non-quantitative data associated with the large datasets by con-verting ordinal, nominal, and categorical variables into numerical variables; normalizing the encoded data to change the values of variables in the dataset to enable a scale applicable to any number of data flux having different origins, without distorting differences in the ranges of values or losing information; performing a grid conversion for representing each variable in the dataset as a gird line to provide a spatial reference to a centroid matrix; converting the centroid matrix into a temporal vector matrix or a colormap matrix by shifting the centroid; and grouping the data associated with the vector matrix or color map matrix to form a plurality of clusters of data having a high co-variance and identifying covariance relationship between various data in the dataset.

公开(公告)号：WO2022272147A1

公开(公告)日：2022-12-29

申请号：PCT/US2022/035019

申请日：2022-06-24

Applicant: THE REGENTS OF THE UNIVERSITY OF CALIFORNIA

Inventor： YELLOWLEES, Peter ,  PARISH, Michelle Burke ,  CHAN, Steven Richard

IPC: G16H10/20 ,  G16H10/60 ,  G16H20/70 ,  G16H50/20 ,  G16H50/50 ,  G06F40/10 ,  G06F40/58 ,  G06N3/04 ,  G06N3/08 ,  G10L15/26 ,  G16H40/67 ,  G10L15/22 ,  G06N20/00 ,  G06N5/01 ,  G06N5/041 ,  G16H30/20

Abstract: Disclosed herein are methods and systems for a training a model for real-time patient diagnosis. A system may include a computer configured to receive audio data and video data of a clinical encounter, the audio data comprising spoken words by an entity and the video data depicting the entity; retrieve clinical data regarding the entity; execute a model using the words of the audio data and the retrieved clinical data regarding the entity as input, the execution causing the model to output a plurality of clinical diagnoses for the entity; concurrently render the corresponding video data and audio data and the plurality of clinical diagnoses via a computing device associated with a user; and store an indication of a selected clinical diagnosis of the plurality of clinical diagnoses responsive to receiving a selection of the clinical diagnosis at the computing device.

公开(公告)号：WO2022216059A1

公开(公告)日：2022-10-13

申请号：PCT/KR2022/004982

申请日：2022-04-06

Applicant: 삼성전자 주식회사

Inventor： 최광용 ,  김미선 ,  서미라 ,  김상헌 ,  정혜순 ,  임연욱

IPC: G10L17/00 ,  G10L15/26 ,  G10L25/48 ,  G06F40/10 ,  G06F40/20

Abstract: 본 문서는 개인화 오디오 정보를 제공하기 위한 전자 장치, 전자 장치에서의 동작 방법 및 비일시적 저장 매체에 관한 것으로서, 일 실 시예에 따르면, 전자 장치는, 오디오 모듈, 메모리 및 상기 오디오 모듈 및 상기 메모리에 전기적으로 연결된 적어도 하나의 프로세서를 포함하며, 상기 적어도 하나의 프로세서는, 사용자 인터랙션 정보를 수신하고, 상기 수신된 사용자 인터랙션 정보를 기반하여 타겟 객체의 특징 정보를 식별하고, 상기 식별된 특징 정보를 분석하여 적어도 하나의 화자를 식별하고, 식별된 적어도 하나의 화자를 기반하여 상기 특징 정보에 대응하는 개인화 오디오 정보를 생성하고, 상기 개인화 오디오 정보를 상기 타겟 객체에 매핑하여 개인화된 콘텐츠를 생성하고, 상기 개인화된 콘텐츠를 저장 및 실행하도록 설정될 수 있다. 다른 실시예도 가능하다.

公开(公告)号：WO2022092487A1

公开(公告)日：2022-05-05

申请号：PCT/KR2021/008778

申请日：2021-07-09

Applicant: 삼성전자주식회사

Inventor： 최형탁 ,  라부루로힛 ,  양해훈

IPC: G06Q10/10 ,  G06F40/10 ,  G06K9/32 ,  G06N20/00 ,  G06N3/08

Abstract: 전자 장치가 개시된다. 전자 장치는 디스플레이, 적어도 하나의 명령을 저장하는 메모리 및 메모리 및 디스플레이와 연결되어 전자 장치를 제어하는 프로세서를 포함하며, 프로세서는 적어도 하나의 명령어를 실행함으로써 디스플레이에 이미지가 디스플레이되는 동안 스케줄 추가를 위한 명령이 입력되면 이미지에 대한 텍스트 인식을 수행하여 복수의 텍스트를 획득하고, 획득된 복수의 텍스트를 제1 신경망 모델에 입력하여 복수의 스케줄 정보 각각에 대응되는 메인 일시 정보 및 메인 일시 정보에 대응되는 서브 일시 정보를 획득하고, 획득된 일시 정보에 기초하여 사용자의 스케줄 정보를 업데이트하며, 제1 신경망 모델은 복수의 일시 정보를 입력받아 메인 일시 정보 및 메인 일시 정보에 대응되는 서브 일시 정보를 출력하도록 학습된다.

公开(公告)号：WO2022085020A1

公开(公告)日：2022-04-28

申请号：PCT/IN2021/050958

申请日：2021-10-06

Applicant: LARSEN & TOUBRO INFOTECH LTD.

Inventor： VYAS, Sachin ,  SALUJA, Satish ,  SARKAR, Jnanendra Prasad ,  KAMAT, Yash

IPC: G06F7/00 ,  G06F16/35 ,  G06F40/10

Abstract: The present invention discloses a method and a system for clustering of short and long text documents. The documents are input through an input module and a pre-processing module overtakes the documents from the input module. The pre-processing module refines the documents and removes unwanted text from the documents. Then a decision driven hybrid text clustering algorithm is applied via different modules to achieve clustering of the documents. Firstly, a context module computes a moment value of a feature signifying at least one feature importance value of the feature for the documents. The moment value is used by a decision module to calculate a decision score. Basis the decision score the documents are split into two sets. A clustering module then forms clusters of the two sets of documents basis n-tuple word distribution. Finally, a convergence module congregates the clusters in a final set of documents.

公开(公告)号：WO2022023988A1

公开(公告)日：2022-02-03

申请号：PCT/IB2021/056838

申请日：2021-07-28

Applicant: WAY2VAT LTD.

Inventor： SIMANTOV, Amos ,  SHILKROT, Roy ,  GAL, Rinon ,  ARDAZI, Shai

IPC: G06F17/10 ,  G06F17/40 ,  G06F40/10 ,  G06F40/279 ,  G06K9/20 ,  G06N3/02

Abstract: Systems and methods for matching document images to tabulated records. Essential features are extracted from document images and converted into a set of document-image vectors. Spreadsheet data is converting a set of sheet-line vectors. A differential parametric algorithm is applied to determine a similarity score metric which is used in a matching algorithm. The matching algorithm is applied to the set of document images and the spreadsheet data.

公开(公告)号：WO2021247610A1

公开(公告)日：2021-12-09

申请号：PCT/US2021/035297

申请日：2021-06-01

Applicant: COGNIZER, INC.

Inventor： PORTER, Jack ,  VELU, Soundararajan ,  MUNIRATHNAM, Vineeth ,  KIRCH, Suzanne ,  BARONIA, Rajiv

IPC: G06F40/30 ,  G06N3/08 ,  G06F40/10 ,  G06F40/20 ,  G06F40/247

Abstract: Semantic frame identification involves associating identified target words in the sentential context of their natural language source with semantic frames from a frame lexical database. The disclosed invention leverages the CapsNet architecture for improved semantic frame identification of a target word in a natural language input. This includes deriving the features of a target word identified in the sentence and extracting the features of the word units and the thematic words around the target word. Through dynamic routing of capsules, the CapsNet is able to filter the candidate frames for the target word to reduce the search space and apply the CapsNet prediction to identify a frame from a frame lexical database.

公开(公告)号：WO2021221535A1

公开(公告)日：2021-11-04

申请号：PCT/RU2020/000696

申请日：2020-12-16

Applicant: ПУБЛИЧНОЕ АКЦИОНЕРНОЕ ОБЩЕСТВО "СБЕРБАНК РОССИИ"

Inventor： ШАВРИНА, Татьяна Олеговна

IPC: G06F40/10 ,  G06N20/00

Abstract: Изобретение относится к области компьютерной техники. Технический результат заключается в обеспечении подбора текстовых данных для аугментации обучающей выборки на основании характеристик текста входной обучающей выборки. Раскрыт компьютерно-реализуемый способ аугментации обучающей выборки для алгоритмов машинного обучения, содержащий этапы, на которых: получают текстовые данные исходной обучающей выборки; выполняют нормализацию данных, при которой выполняется разделение текста на предложения и очистка текста от символов; выполняют векторизацию нормализованных предложений, при этом в ходе упомянутого преобразования осуществляется: разбиение каждого полученного предложения на минимально значимые части, представляющие собой слова и знаки препинания (токенизация); формирование векторных представлений для каждого нормализованного текста на основании входящих в него токенов (значимых частей); формируют текстовый индекс по векторным представлениям текстовых данных, при этом текстовый индекс формируется из векторного пространства, формируемого из текстов, расположенных в открытых источниках, и метаданных; осуществляют аугментацию исходной обучающей выборки с помощью подбора релевантных векторных представлений текстов на основании определения меры близости в векторном пространстве на основании поискового индекса.

公开(公告)号：WO2021119246A1

公开(公告)日：2021-06-17

申请号：PCT/US2020/064197

申请日：2020-12-10

Applicant: TINYIVY, INC.

Inventor： SILVERZWEIG, Zachary

IPC: G06F40/00 ,  G06F40/10 ,  G06F40/126

Abstract: An unambiguous phonics system (UPS) is capable of presenting text in a format with unambiguous pronunciation. The system can translate input text written in a given language (e.g., English) into a UPS representation of the text written in a UPS alphabet. A unique UPS grapheme can be used to represent each unique grapheme-phoneme combination in the input text. Thus, each letter of the input text is represented in the UPS spelling and each letter of the UPS spelling unambiguously indicates the phoneme used. For all the various grapheme-phoneme combinations for a given input grapheme, the corresponding UPS graphemes can be constructed to have visual similarity with the given input grapheme, thus easing an eventual transition from UPS spelling to traditional spelling. The UPS can include translation, complexity scoring, word/phoneme-grapheme searching, and other module. The UPS can also include techniques to provide efficient, level-based training of the UPS alphabet.