Natural Language Patents (Class 704/9)
  • Patent number: 11031139
    Abstract: Systems and methods are provided for presenting aggregate data in response to a natural language user input. In one example, a system includes a display and a computing device coupled to the display and storing instructions executable to receive a natural language user input, process the natural language user input, in response to determining that the user input includes a request to display two different plots of record data specific to the subject, generate, with the virtual assistant, a single graph including the two different plots of record data based on the processed natural language user input, the two different plots of record data plotted from two different record data sets, one or more aspects of the single graph selected based on an overlapping parameter for each of the two different record data sets, and output, to the display, the single graph as part of a communication thread.
    Type: Grant
    Filed: February 8, 2019
    Date of Patent: June 8, 2021
    Assignee: General Electric Company
    Inventors: Omer Barkol, Renato Keshet, Andreas Tzanetakis, Constance Anne Rathke, Reuth Goldstein, Michelle Townshend
  • Patent number: 11032222
    Abstract: Techniques are described that determine when content to be shared by a user may be offensive, and providing the user with a notification that the content may be offensive. In some examples, the user may be provided with a notification that includes a selectable option allowing the user to withdraw or undo the content from being shared, additional information regarding why the content was determined to be offensive, content sharing guidelines, and/or advice for how to revise the content to be less offensive. In some examples, the notification may be presented for a predetermined period of time, and the content may be held in a pending state and not shared with other users until expiration of the predetermined period.
    Type: Grant
    Filed: August 22, 2019
    Date of Patent: June 8, 2021
    Assignee: Facebook, Inc.
    Inventors: Francesco Fogu, Mindi Yuan, Yoav Shapira, Alana Conner
  • Patent number: 11030517
    Abstract: A summary obtaining method includes obtaining a target document from which a summary is to be obtained; dividing the target document into sentences and dividing each sentence into words; and obtaining a classifier constructed based on a long short-term memory (LSTM) model. The classifier comprises a word-level LSTM layer configured to determine a vector of each sentence according to the words in each sentence, and a sentence-level LSTM layer configured to determine a vector of the target document according to the vector of each sentence. The method further includes determining, by using the classifier, a probability that each sentence is a summary sentence, according to the vector of the target document and a vector of the sentence; and determining a sentence with the probability satisfying a predetermined condition for a summary sentence, as a summary sentence of the target document, and forming the summary according to the summary sentence.
    Type: Grant
    Filed: October 14, 2019
    Date of Patent: June 8, 2021
    Assignee: TENCENT TECHNOLOGY (SHENZHEN) COMPANY LIMITED
    Inventors: Xuxiang Wang, Dan Nie
  • Patent number: 11030405
    Abstract: A method and a device for generating a statement are provided. An embodiment includes obtaining an input statement by a user over a terminal; determining at least one target position in the input statement; according to a target position in the at least one target position, determining a statement to be inserted from the input statement; based on the statement to be inserted, determining at least one candidate word to be inserted in the target position; inserting a candidate word in the at least one candidate word into the target position to obtain a candidate statement; and sending the candidate statement to the terminal of the user for display. The embodiment can make candidate statements displayed on the terminal of the user more diversified to assist the user to select an expected statement from candidate statements, thereby saving time of the user spending on manually inserting words.
    Type: Grant
    Filed: September 15, 2020
    Date of Patent: June 8, 2021
    Assignee: BEIJING BYTEDANCE NETWORK TECHNOLOGY CO., LTD.
    Inventor: Xiaoshuai Chen
  • Patent number: 11031003
    Abstract: Technology is disclosed for providing dynamic identification and extraction or tagging of contextually-coherent text blocks from an electronic document. In an embodiment, an electronic document may be parsed into a plurality of content tokens that each corresponds to a portion of the electronic document, such as a sentence or a paragraph. Employing a sliding window approach, a number of token groups are independently analyzed, where each group of tokens has a different number of tokens included therein. Each token group is analyzed to determine confidence scores for various determinable contexts based on content included in the token set. The confidence scores can then be processed for each token group to determine an entropy score for the token group. In this way, one of the analyzed token groups can be selected as a representative text block that corresponds to one of the plurality of determinable contexts.
    Type: Grant
    Filed: May 25, 2018
    Date of Patent: June 8, 2021
    Assignee: MICROSOFT TECHNOLOGY LICENSING, LLC
    Inventors: Abedelkader Asi, Liron Izhaki-Allerhand, Ran Mizrachi, Royi Ronen, Ohad Jassin
  • Patent number: 11030415
    Abstract: A document analysis system trains a document embedding model configured to receive a set of word embeddings for an ordered set of words in a document and generate a document embedding for the document. The document embedding is a representation of the document in a latent space that characterizes the document with respect to properties such as structure, content, and sentiment. The document embedding may represent a prediction of a set of words that follow the last word in the ordered set of words of the document. The document embedding model may be associated with a convolutional neural network (CNN) architecture that includes one or more convolutional layers. The CNN architecture of the document embedding model allows the document analysis system to overcome various difficulties of existing document embedding models, and allows the document analysis system to easily process variable-length documents that include a variable number of words.
    Type: Grant
    Filed: June 5, 2019
    Date of Patent: June 8, 2021
    Assignee: The Toronto-Dominion Bank
    Inventors: Maksims Volkovs, Tomi Johan Poutanen
  • Patent number: 11023477
    Abstract: A method for performing a fuzzy search in encrypted data includes receiving an encrypted search token corresponding to a search keyword with an untrusted server computing device and generating inner product values based on a function-hiding inner product encryption operation of the encrypted query vectors encrypted node vectors in an encrypted tree stored in the memory of the untrusted server computing device. The method further includes transmitting, with the untrusted server computing device, the encrypted keyword stored in the leaf node to a client computing device in response to the first inner product value exceeding a first predetermined similarity threshold corresponding to a similarity of the first query vector to the leaf node vector, the fuzzy search not revealing plaintext contents for any of a keyword stored in the leaf node, the search keyword, or a fuzziness parameter.
    Type: Grant
    Filed: December 27, 2017
    Date of Patent: June 1, 2021
    Assignee: Robert Bosch GmbH
    Inventors: Xinxin Fan, Qingji Zheng
  • Patent number: 11024406
    Abstract: Systems and methods for analyzing a medical report to determine whether the medical report includes at least one instance of at least one category selected from a group consisting of: gender error, laterality error, and critical finding. In some embodiments, one or more portions of text are identified from the medical report. Contextual information associated with the medical report is used to determine whether the identified one or more portions of text comprise at least one instance of at least one category selected from the group.
    Type: Grant
    Filed: March 12, 2013
    Date of Patent: June 1, 2021
    Assignee: Nuance Communications, Inc.
    Inventors: Sepehr Sadeghi, Atul T. Mistry
  • Patent number: 11023418
    Abstract: The present invention provides a data management system. More specifically, the present invention relates to a keyword-based data management system and method for managing and utilizing multiple pieces of data generated when application programs executed in various types of computing apparatuses are used, and improving a management environment. Proposed according to an embodiment of the present invention is a new procedure for selecting a keyword-of-interest storage before an operation of using a computing apparatus, by classifying data used or collected in the past and storing the same in a keyword-of-interest-based storage. Accordingly, a user can easily access collected data associated with his/her interest, which leads the user to systematically manage data in which the user is interested, and delete and dispose of data in which the user is not interested.
    Type: Grant
    Filed: February 7, 2018
    Date of Patent: June 1, 2021
    Inventor: Jin Weon Kim
  • Patent number: 11023662
    Abstract: Systems and methods for providing adaptive surface texture in auto-drafted patent documents are disclosed. Exemplary implementations may: obtain a plurality of style attributes, a given style attribute being descriptive of a type of document characteristic that is often-repeated or habitual in patent documents; obtain values for the plurality of style attributes, a given value of a style attribute being descriptive of a specific document characteristic that is often-repeated or habitual in one or more sample patent documents; and provide an auto-drafted patent document exhibiting the values for the plurality of style attributes.
    Type: Grant
    Filed: April 3, 2020
    Date of Patent: June 1, 2021
    Assignee: Specifio, Inc.
    Inventors: Ian C. Schick, Kevin Knight, Jay Priyadarshi, Xing Shi
  • Patent number: 11023654
    Abstract: A method for generating an appendix from document content is provided. Included is analyzing a structure of a document to identify at least one logical element in a content of a page of the document; identifying at least one relation within the identified at least one logical element to determine an appendix topic; determining a degree of relevance of the at least one relation identifying the appendix topic; extracting an appendix content from at least one source document, based on determining a relevance of the source document to the appendix topic; and formatting the appendix based on the extracted appendix content.
    Type: Grant
    Filed: February 21, 2020
    Date of Patent: June 1, 2021
    Assignee: International Business Machines Corporation
    Inventors: Anand B. Bhagwat, Amit P. Bohra, Krishna Kummamuru, Abhishek Shivkumar
  • Patent number: 11023681
    Abstract: Embodiments for co-reference resolution and entity linking from unstructured text data by a processor. Semantic co-references and mentions of one or more entities may be resolved occurring in unstructured text data by linking the one or more entities using a domain knowledge ontology.
    Type: Grant
    Filed: September 19, 2018
    Date of Patent: June 1, 2021
    Assignee: INTERNATIONAL BUSINESS MACHINES CORPORATION
    Inventors: Yassine Lassoued, Lea Deleris, Stephane Deparis, Killian Levacher, Charles A. Jochim, Yufang Hou, Elizabeth Daly
  • Patent number: 11023679
    Abstract: An apparatus for automatically mapping a verbatim narrative to a term in a medical terminology dictionary includes a natural language processor and a comparator. The natural language processor processes terms from the medical terminology dictionary and from a medical coding decision database to generate a processed database that also includes the original terms from the medical terminology dictionary and the medical coding decision database. The natural language processor also processes the verbatim narrative. The comparator compares the processed verbatim narrative to the terms in the processed database and determines whether the processed verbatim narrative is an exact match to a term in the processed database. The verbatim narrative is mapped to the term in the medical terminology dictionary that corresponds to the term in the processed database that is an exact match. The verbatim narratives may include adverse event narratives, concomitant medication narratives, or other types of narratives.
    Type: Grant
    Filed: February 27, 2017
    Date of Patent: June 1, 2021
    Assignee: Medidata Solutions, Inc.
    Inventors: Patricia Allen, Andrew Howland, Philip Beineke, Mark Chandler, Michael Elashoff, Mladen Laudanovic, Jingshu Liu, Michael Cestone, Jenny Liu
  • Patent number: 11024424
    Abstract: According to some aspects, a system for automatically processing text comprising information regarding a patient encounter to assign medical codes to the text is provided. The system comprises at least one storage medium storing processor-executable instructions, and at least one processor configured to execute the processor-executable instructions to perform analyzing the text to extract a plurality of facts from the text, identifying at least one of the plurality of facts to be excluded from consideration when assigning medical codes to the text, and evaluating each of the plurality of facts, except for the identified at least one fact, to assign one or more medical codes to the text.
    Type: Grant
    Filed: October 27, 2017
    Date of Patent: June 1, 2021
    Assignee: Nuance Communications, Inc.
    Inventors: Weiyi Sun, Ravi Kondadadi, Brian William Delaney, Girija Yegnanarayanan
  • Patent number: 11019107
    Abstract: Some aspects of the present disclosure relate to systems and methods for identifying potential violation conditions from electronic communications. In one embodiment, a method includes receiving data associated with an electronic communication and detecting, from the received data, and using a trainable model, an indicator of a potential violation condition, where the violation condition is associated with an activity that is a violation of a predetermined standard. The method also includes, responsive to detecting the indicator of the potential violation condition, marking the electronic communication as being associated with a potential violation condition, and presenting the potential violation condition to a user for review.
    Type: Grant
    Filed: February 19, 2018
    Date of Patent: May 25, 2021
    Assignee: Digital Reasoning Systems, Inc.
    Inventors: John Wagster, Robert Metcalf, Keith Ellis Massey, Kenneth Loran Graham, Sarah Cannon, Adam Jaggers, Vishnuvardhan Balluru, Bill Dipietro
  • Patent number: 11017182
    Abstract: A translation device includes a character string output unit, a fixed phrase acquisition unit, a matching unit, a division processor, and a translation result output unit. The character string output unit outputs a plurality of character strings each configuring at least part of text in a first language and including a head character of the text, in order of character string length. The fixed phrase acquisition unit acquires a fixed phrase in the first language associated with a translation phrase in a second language. The matching unit calculates a similarity between each of the character strings and the fixed phrase in order of the character string length. The division processor divides the text at a position achieving a maximum similarity among values of the similarity exceeding a first threshold. The translation result output unit outputs the translation phrase as translation of the text thus divided.
    Type: Grant
    Filed: February 26, 2019
    Date of Patent: May 25, 2021
    Assignee: PANASONIC INTELLECTUAL PROPERTY MANAGEMENT CO., LTD.
    Inventor: Natsuki Saeki
  • Patent number: 11017765
    Abstract: A method for use with a computing device is provided. The method may include executing one or more programs of an intelligent digital assistant system at a processor and presenting a user interface to a user. At the processor, the method may include receiving natural language user input from the user, parsing the user input at an intent handler to determine an intent template with slots, populating the slots in the intent template with information from user input, and performing resolution on the intent template to partially resolve unresolved information. If a slot with missing slot information exists in the partially resolved intent template, a loop may be executed at the processor to fill the slots. The method may include, at the processor, determining that all required information is available and resolved and generating a rule based upon the intent template with all required information being available and resolved.
    Type: Grant
    Filed: December 2, 2019
    Date of Patent: May 25, 2021
    Assignee: Microsoft Technology Licensing, LLC
    Inventors: Oz Solomon, Christopher Brian Quirk, Han Yee Mimi Fung, Keith Coleman Herold
  • Patent number: 11016964
    Abstract: The selection of search results to display can be based at least in part upon an intent determined for a search query. Various intent terms can be used to specify intent or context for a search query in order to obtain a certain type of result. The intent terms can be suggested or provided to a user in such a way as to train the user as to their existence and usage. A submitted search query is analyzed to detect the intent term and determine the overall intent of the query. One or more targeted intent queries are executed against determined data sources to obtain results relevant to that intent, such as new content items for a particular topic or items that are similar to those offered by a particular store. A diverse assortment of results can then be presented to the user.
    Type: Grant
    Filed: September 12, 2017
    Date of Patent: May 25, 2021
    Assignee: AMAZON TECHNOLOGIES, INC.
    Inventors: Lisa Jane Hinegardner, Andrew Dennis Hamel, Srikanth Thirumalai
  • Patent number: 11017777
    Abstract: A language processing method and apparatus is disclosed. A language processing apparatus using a neural network may obtain context information from a source text using a neural network-based encoder, generate a prefix token from the context information using a neural network-based main decoder, generate a token sequence including at least two successive tokens sequentially following the prefix token using a skip model in response to the prefix token satisfying a preset condition, and indicate a target text in which the prefix token and the token sequence are combined as an inference result with respect to the source text.
    Type: Grant
    Filed: August 22, 2018
    Date of Patent: May 25, 2021
    Assignee: Samsung Electronics Co., Ltd.
    Inventors: Min-Joong Lee, Hodong Lee
  • Patent number: 11017307
    Abstract: A method of generating text having related purposes using a generative adversarial network (GAN) includes receiving a limited dataset including real data with related cognitive value types (types). The method includes applying loss functions to portions of the real data. The portions of the real data are each identified as having one of the types. The loss functions ensure alignment of the portions with corresponding types. The method includes embedding the real data into an encoder output that includes an embedded vector for the cognitive value types. The method includes generating an additional parameter set supplemental to the real data and configured to enhance an expressiveness of a model. The method includes generating statements based on the additional parameter set and the encoder output. The statements include a style of one of the cognitive value types and are related to a common issue addressed by the GAN.
    Type: Grant
    Filed: February 18, 2019
    Date of Patent: May 25, 2021
    Assignee: FUJITSU LIMITED
    Inventors: Pouya Pezeshkpour, Ramya Malur Srinivasan, Ajay Chander
  • Patent number: 11018890
    Abstract: A building system including one or more memory devices configured to store instructions thereon, the instructions causing one or more processors to generate a temporary relationship between a first entity and a second entity of a space graph, cause the space graph to include the temporary relationship, perform one or more control operations based on the space graph including the temporary edge, receive new building data from the one or more building data sources, determine whether to generate a permanent relationship to replace the temporary relationship based on the new building data, and update the space graph by causing the permanent relationship to replace the temporary relationship of the space graph in response to a determination to generate the permanent relationship to replace the temporary relationship by causing a permanent edge to replace the temporary edge.
    Type: Grant
    Filed: June 8, 2020
    Date of Patent: May 25, 2021
    Assignee: Johnson Controls Technology Company
    Inventors: Youngchoon Park, Sudhi Sinha
  • Patent number: 11019015
    Abstract: Techniques are described that determine when content to be shared by a user may be offensive, and providing the user with a notification that the content may be offensive. In some examples, the user may be provided with a notification that includes a selectable option allowing the user to withdraw or undo the content from being shared, additional information regarding why the content was determined to be offensive, content sharing guidelines, and/or advice for how to revise the content to be less offensive. In some examples, the notification may be presented for a predetermined period of time, and the content may be held in a pending state and not shared with other users until expiration of the predetermined period.
    Type: Grant
    Filed: August 22, 2019
    Date of Patent: May 25, 2021
    Assignee: Facebook, Inc.
    Inventors: Yoav Shapira, Alana Conner, Mindi Yuan, Francesco Fogu, Samuel James Parker, Fan Yang
  • Patent number: 11010560
    Abstract: Methods and systems for natural language processing include generating respective feature vectors, for each word in an input, based on sequences of input words of different respective lengths. The respective feature vectors for each word in the input are combined to form a combined vector for each word. A hidden state is determined for each word in the input based on the combined vector. The hidden states for all words in the input are combined to form a multi-resolution context vector. A natural language processing action is performed using the multi-resolution context vector.
    Type: Grant
    Filed: November 8, 2018
    Date of Patent: May 18, 2021
    Assignee: INTERNATIONAL BUSINESS MACHINES CORPORATION
    Inventors: Song Xu, Zhili Guo, HongLei Guo, Shiwan Zhao, Pitipong Jun Sen Lin, Elaine Branagh, Zhong Su
  • Patent number: 11010561
    Abstract: Techniques for predicting sentiment from textual data are described herein. In some examples, the described techniques utilize a sentiment prediction model having bidirectional long short-term memory (LSTM) networks with one or more convolution-and-pooling stages. The bidirectional LSTM networks process vector representations of words in a textual word sequence to determine forward and backward word-level context feature vectors. Forward and backward phrase-level feature vectors are determined based on the forward and backward word-level context feature vectors. The one or more convolution-and-pooling stages pool the forward and backward phrase-level feature vectors to determine pooled phrase-level feature vectors. A sentiment representing the textual word sequence is determined based on the pooled phrase-level feature vectors.
    Type: Grant
    Filed: December 27, 2018
    Date of Patent: May 18, 2021
    Assignee: Apple Inc.
    Inventor: Jerome R. Bellegarda
  • Patent number: 11010545
    Abstract: A method for narrating a table using at least one narration template, wherein the table is extracted from a data source is provided. The method may also include performing structural analysis on the extracted table. The method may further include selecting a structural template based on the structural analysis of the extracted table. Additionally, the method may include selecting a narration template based on the selected structural template. The method may further include narrating the extracted table based on the selected narration template and displaying a narrative output of the extracted table.
    Type: Grant
    Filed: April 24, 2019
    Date of Patent: May 18, 2021
    Assignee: International Business Machines Corporation
    Inventors: Chinnappa Guggilla, Ashish Mungi, Purushothaman K. Narayanan, Ankur S. Parikh, Krishma Singla, Bijo A. Thomas
  • Patent number: 11011156
    Abstract: A computer-implemented method for training a model is disclosed. The model is capable of retaining a history of one or more preceding elements and has a direction of prediction. The method includes obtaining a training sequence of elements. The method also includes splitting the training sequence into a plurality of parts. The method further includes selecting one part of the plurality of the parts depending on the direction of the model to generate a modified training data. The method includes further training the model using the modified training data.
    Type: Grant
    Filed: April 11, 2019
    Date of Patent: May 18, 2021
    Assignee: INTERNATIONAL BUSINESS MACHINES CORPORATION
    Inventor: Gakuto Kurata
  • Patent number: 11010360
    Abstract: A method for extending a tag of an information resource in an information network may be provided. The method may include providing a group of tags, stored in a database, in a user interface, and selecting at least one tag out of the group of tags and linking the tag to an information resource using the user interface. Furthermore, the method may include extending the tag, such that the resulting extended tag may be a semantically structured string of characters building a first expression including a first operand, a first operator and a first value, wherein the tag may be the operand, such that the extended tag is machine-interpretable by a parser, such that the information resource may be tagged with the machine-interpretable, extended tag.
    Type: Grant
    Filed: May 17, 2018
    Date of Patent: May 18, 2021
    Assignee: International Business Machines Corporation
    Inventors: Matthias Falkenberg, Michael Junginger, Andreas Nauerz, Thomas Steinheber
  • Patent number: 11010180
    Abstract: The present disclosure discloses method and guidance providing system for providing real-time guidance to users during troubleshooting of devices. The guidance providing system receives user input indicating troubleshooting actions on device, by user, where user input comprises textual input and non-textual input, identifies object in device associated with troubleshooting actions and state of user while performing troubleshooting actions, from non-textual input. The object identified from non-textual input is mapped with textual input to obtain mapped user input. Consistency of mapped user input is verified based on conversation information and domain specific content associated with device. Sequence of instructions corresponding to mapped user input, based on state of user and one or more similar troubleshooting actions, may be extracted from knowledge database.
    Type: Grant
    Filed: July 17, 2018
    Date of Patent: May 18, 2021
    Assignee: Wipro Limited
    Inventors: Manjunath Ramachandra Iyer, Meenakshi Sundaram Murugeshan
  • Patent number: 11010565
    Abstract: A system includes a natural language processor to drive a financial transaction session with a chat application without engaging a context controller until such time that a completion threshold for a context state of the financial transaction session is met. Once the context state meets the completion threshold, the natural language processor engages the context controller with the context state, and the context controller applies the context state from the natural language processor to drive the financial transaction session between the chat application and a backend server system.
    Type: Grant
    Filed: January 4, 2021
    Date of Patent: May 18, 2021
    Assignee: Globant España, S.A.
    Inventor: Martin Migoya
  • Patent number: 11010492
    Abstract: Methods and systems are provided for detecting privacy leakage risks in text. One example method generally includes receiving, at a computing device from a client device, a keyword and generating, by the computing device, a combined query comprising the keyword and a sensitive topic query associated with a sensitive topic. The method further includes transmitting the keyword from the computing device to a search engine and transmitting the combined query from the computing device to the search engine. The method further includes receiving, at the computing device from the search engine, a number of search results for the keyword and receiving, at the computing device from the search engine, a number of search results for the combined query. The method further includes determining, by the computing device, a confidence score and transmitting the confidence score from the computing device to the client device.
    Type: Grant
    Filed: September 10, 2018
    Date of Patent: May 18, 2021
    Assignee: NortonLifeLock Inc.
    Inventors: Ashwin Kayyoor, Petros Efstathopoulos
  • Patent number: 11010179
    Abstract: In one embodiment, a method includes receiving a user input by the first user from a client system associated with a first user, parsing the user input to identify one or more n-grams associated with the user input, accessing a user profile associated with the first user, wherein the user profile is stored in a first data store, accessing ontology data based on the one or more identified n-grams from one or more information graphs, wherein the one or more information graphs are stored in one or more second data stores, respectively, determining contextual information associated with the user input, generating semantic information by aggregating the user profile, ontology data, and contextual information, generating a feature representation for the identified one or more n-grams based on the semantic information, and resolving one or more entities associated with the one or more n-grams based on the feature representation.
    Type: Grant
    Filed: April 30, 2018
    Date of Patent: May 18, 2021
    Assignee: Facebook, Inc.
    Inventors: Vivek Natarajan, Baiyang Liu, Xiaohu Liu, Ahmed Aly
  • Patent number: 11010546
    Abstract: A method for narrating a table using at least one narration template, wherein the table is extracted from a data source is provided. The method may also include performing structural analysis on the extracted table. The method may further include selecting a structural template based on the structural analysis of the extracted table. Additionally, the method may include selecting a narration template based on the selected structural template. The method may further include narrating the extracted table based on the selected narration template and displaying a narrative output of the extracted table.
    Type: Grant
    Filed: April 24, 2019
    Date of Patent: May 18, 2021
    Assignee: International Business Machines Corporation
    Inventors: Chinnappa Guggilla, Ashish Mungi, Purushothaman K. Narayanan, Ankur S. Parikh, Krishma Singla, Bijo A. Thomas
  • Patent number: 11010553
    Abstract: A computer implemented method is provided for recommending at least one author of a plurality of authors to a user based on a learned lexicon of the user. First communication content pertaining to a user is obtained. Second communication content pertaining to an author is obtained. A lexicon of the user is learned from the first communication content to identify a set of words known to the user. High frequency words are identified from the second communication content of the author, the high frequency words being one or more words having an occurrence frequency in the second communication content that is higher than a first predetermined threshold occurrence frequency level. The author is suggested to the user upon detecting that the high frequency words are not in the lexicon of the user.
    Type: Grant
    Filed: April 18, 2018
    Date of Patent: May 18, 2021
    Assignee: INTERNATIONAL BUSINESS MACHINES CORPORATION
    Inventors: Pasquale A. Catalano, John S. Werner, Andrew G. Crimmins, Arkadiy O. Tsfasman
  • Patent number: 11010838
    Abstract: One embodiment can provide a system for detecting optimizing a damage detection result. During operation, the system can obtain a digital image of a damaged vehicle, identify a set of candidate damaged areas from the digital image as the damage detection result. The system can then extract a set of feature vectors corresponding to the set of candidate damaged areas For each candidate damaged area, the system can calculate a set of similarity features between the candidate damaged area and other candidate damaged areas in the set of candidate damaged areas based on the set of feature vectors. The system can input the set of similarity features to a damage prediction module. The system can then determine whether the candidate damaged area is an exceptional area based on an output of the damage prediction module to optimize the damage detection result.
    Type: Grant
    Filed: August 26, 2019
    Date of Patent: May 18, 2021
    Assignee: Advanced New Technologies Co., Ltd.
    Inventor: Juan Xu
  • Patent number: 11011161
    Abstract: A computer-implemented method is provided for generating a plurality of templates. The method includes obtaining, by a processor device, a Recurrent Neural Network Language Model (RNNLM) trained using a first set of text data. The method further includes adapting, by the processor device, the RNNLM using a second set of text data by adding a new node corresponding to a class in both an input layer and an output layer of the RNNLM, the class being obtained from the second set of text data. The method also includes generating, by the processor device, the plurality of templates using the adapted RNNLM.
    Type: Grant
    Filed: December 10, 2018
    Date of Patent: May 18, 2021
    Assignee: INTERNATIONAL BUSINESS MACHINES CORPORATION
    Inventors: Masayuki Suzuki, Toru Nagano, Nobuyasu Itoh, Gakuto Kurata
  • Patent number: 11003854
    Abstract: A method retrains a lexical analysis model. One or more processors divide a token in a document into a first sub-token having a boundary that begins with a start point of a mention and a second sub-token that ends at an end point of the mention. The processor(s) determine that the start/end point of the mention does not coincide with a start/end point of the first sub-token; concatenate the first sub-token with the second sub-token to create a concatenated token; and store the concatenated token in a user dictionary for the document. The processor(s) perform a morphological analysis of the document using the concatenated token from the user dictionary, and retrain a lexical analysis model for the document by leveraging the performing of the morphological analysis of the document.
    Type: Grant
    Filed: October 30, 2018
    Date of Patent: May 11, 2021
    Assignee: International Business Machines Corporation
    Inventors: Ken Kumagai, Akihiro Nakayama, Kei Sugano
  • Patent number: 11003849
    Abstract: Technologies for natural language processing include a computing device that loads natural language text data that includes multiple words, such as a naturally elicited response or comment data. The computing device applies a constituency parser to the natural language text data to generate a constituency parse tree. The constituency parse tree may include multiple nodes that each correspond to a hierarchical constituent of the natural language text. The computing device applies a constituency rule to identify clauses in the natural language text data. The constituency rule may identify a left-most daughter of each subordinate clause node of the constituency parse tree. Each subordinate clause is distinct, preventing dependencies from crossing clause boundaries. The computing device applies a dependency parser to each clause to generate a corresponding dependency parse.
    Type: Grant
    Filed: September 7, 2018
    Date of Patent: May 11, 2021
    Assignee: Press Ganey Associates, LLC
    Inventors: Zach W. Childers, Kyle Robertson, Taylor Turpen
  • Patent number: 11003855
    Abstract: A method includes generating entigen groups from phrases of a related topic, where the entigen groups represents most likely meanings of the phrases. The method further includes identifying a first entigen group and a second entigen group, where a first meaning of the first entigen group is similar to, but not identical to, a second meaning of the second entigen group. The method further includes determining a meaning difference between the first meaning and the second meaning. When the first meaning has priority over the second meaning, the method further includes updating the second entigen group with the meaning difference to produce an updated second entigen group. When the second meaning has priority over the first meaning, the method further includes updating the first entigen group with the meaning difference to produce an updated first entigen group.
    Type: Grant
    Filed: March 21, 2019
    Date of Patent: May 11, 2021
    Assignee: entigenlogic LLC
    Inventor: Gary W. Grube
  • Patent number: 11004447
    Abstract: A speech processing apparatus includes a speech processor for acquiring information on a relationship between a first user and a second user on the basis of a spoken language of the first user, a storage for storing a first user database (DB) that includes personal information of the second user identified by a second identification tag and a second user DB that includes personal information of the first user identified by a first identification tag, and a DB managing processor for updating the first user DB and the second user DB based on the acquired information on the relationship between the first user and the second user.
    Type: Grant
    Filed: July 13, 2017
    Date of Patent: May 11, 2021
    Assignees: Hyundai Motor Company, Kia Motors Corporation
    Inventors: Jin Ho Lee, Jae Min Joh
  • Patent number: 11003705
    Abstract: A system, computer program product, and method are provided to leverage a taxonomy service to format ground truth data. An artificial intelligence platform processes ground truth data, including identification of one or more applicable taxonomy tags. The identified tags are filtered and applied to the ground truth data, thereby constructing an output string that incorporates the ground truth data together with one or more of the identified tags, effectively transforming the ground truth data. Application of the transformed ground truth data is employed to accurately identify the source and/or meaning of the natural language, and in one embodiment, to produce a physical action or transformation of a physical hardware device.
    Type: Grant
    Filed: February 26, 2018
    Date of Patent: May 11, 2021
    Assignee: International Business Machines Corporation
    Inventors: Stefan A. Van Der Stockt, Sihang B. Fang, Sarah Lynch, Joseph L. Sharpe, III, Rahul P. Akolkar, Brian E. Bissell, Manali J. Chanchlani
  • Patent number: 11000236
    Abstract: Infusion devices and related medical devices, patient data management systems, and methods are provided for monitoring a physiological condition of a patient. An exemplary method of querying a database involves receiving an input query from a client device, identifying a logical layer of a plurality of different logical layers of the database for searching based at least in part on the input query, generating a query statement for searching the identified logical layer of the plurality of different logical layers of the database based at least in part on the input query, querying the identified logical layer of the database using the query statement to obtain result data, and providing a search result influenced by the result data.
    Type: Grant
    Filed: March 22, 2018
    Date of Patent: May 11, 2021
    Assignee: Medtronic MiniMed, Inc.
    Inventors: Yuxiang Zhong, Pratik Agrawal, Huzefa F. Neemuchwala, Sinu Bessy Abraham, Boyi Jiang, Michael P. Stone, Kevin E. Velado, Chantal M. McMahon, Siddharth Arunachalam, Anupam Phukan
  • Patent number: 11003718
    Abstract: A system can enable a global search of security data of a client base. The system can include a processor operable to record anonymity values set by clients of the client base, and to receive search requests including one or more search parameters from the clients. Upon receipt of a search request, processor can generate a result set for the received search request and determine an aggregated anonymity value for the result set. The processor further may compare the aggregated anonymity value of the results set with a set anonymity value for each of the clients for filtering or removing the data points or information of the one or more clients with the set anonymity value that is greater than the aggregate anonymity value from the result set.
    Type: Grant
    Filed: June 12, 2018
    Date of Patent: May 11, 2021
    Assignee: Secureworks Corp.
    Inventors: Lewis McLean, Paul Stansell
  • Patent number: 11003993
    Abstract: This document generally describes a neural network training system, including one or more computers, that trains a recurrent neural network (RNN) to receive an input, e.g., an input sequence, and to generate a sequence of outputs from the input sequence. In some implementations, training can include, for each position after an initial position in a training target sequence, selecting a preceding output of the RNN to provide as input to the RNN at the position, including determining whether to select as the preceding output (i) a true output in a preceding position in the output order or (ii) a value derived from an output of the RNN for the preceding position in an output order generated in accordance with current values of the parameters of the recurrent neural network.
    Type: Grant
    Filed: December 9, 2019
    Date of Patent: May 11, 2021
    Assignee: Google LLC
    Inventors: Samy Bengio, Oriol Vinyals, Navdeep Jaitly, Noam M. Shazeer
  • Patent number: 10997224
    Abstract: A computer implemented method and system for detecting a profanity level of a 140-character textual post generated by a user within one or more communities includes: identifying all words present in the 140-character textual post; assigning a weight to each identified word based on whether or not each identified word occurs in one or more databases; calculating a net profanity factor for each identified word based on the assigned weight; calculating a sum of net profanity factors of all the identified words; and publishing or not publishing the 140-character textual post based on the sum of net profanity factors.
    Type: Grant
    Filed: November 26, 2017
    Date of Patent: May 4, 2021
    Assignee: SKOPIC, INC.
    Inventors: Murali P. Pidathala, Ravindra B. Gudapati, Suneel Nulu
  • Patent number: 10997227
    Abstract: Systems and methods are disclosed herein for processing a natural language query on data tables. According to some embodiments, a natural language query may be originated by a user via a user interface. The natural language query may be parsed to obtain a query term, and a grid range may be identified in a data table as relevant to the query term. A table summary may be prepared including a plurality of data entities based on the grid range. A logic operation may then be determined to apply on the plurality of data entities to derive the query term. The logic operation may then be translated into a formula executable on the data table, and the formula is applied on the data table to generate a result in response to the natural language query.
    Type: Grant
    Filed: January 18, 2017
    Date of Patent: May 4, 2021
    Assignee: Google LLC
    Inventors: Nikunj Agrawal, Mukund Sundararajan, Shrikant Ravindra Shanbhag, Kedar Dhamdhere, Garima, Kevin Snow McCurley, Rohit Ananthakrishna, Daniel Adam Gundrum, Juyun June Song, Rifat Ralfi Nahmias
  • Patent number: 10997371
    Abstract: A response system which realizes an automatic chatting response while reducing investment to a computer necessary for a learning process in artificial intelligence is provided. A chat pattern generation unit of a terminal device generates information on a chat pattern which is a pattern of input fragments which are inputs in a chat and response fragments which are responses to the input fragments. A learning processing unit performs a learning process on the relationship between input text corresponding to the input fragments and the response fragments on the basis of the generated chat pattern and outputs input interpretation knowledge data acquired as a result of the learning process. A response knowledge data transmission unit transmits response knowledge data based on response fragments included in the chat pattern generated by the chat pattern generation unit to an automatic response server device.
    Type: Grant
    Filed: June 7, 2018
    Date of Patent: May 4, 2021
    Assignee: JE International Corporation
    Inventor: Minsu Kim
  • Patent number: 10997404
    Abstract: A system configured to receive a draft patent specification, provide the draft patent specification to a first neural network (NN), classify, using the first NN, paragraphs of the draft patent specification, provide the one or more paragraphs to a second NN, generate, via the second NN, one or more logical components based on the one or more paragraphs, provide the one or more logical components to a generative adversarial network (GAN), generate, via the GAN, a plurality of image components based on the one or more logical components, provide the plurality of image components to a computing device for modification.
    Type: Grant
    Filed: July 18, 2019
    Date of Patent: May 4, 2021
    Assignee: CAPITAL ONE SERVICES, LLC
    Inventors: Jeremy Edward Goodsitt, Austin Walters, Anh Truong, Vincent Pham, Fardin Abdi Taghi Abad, Mark Watson, Reza Farivar, Kenneth Taylor
  • Patent number: 10996925
    Abstract: Implementations set forth herein relate an automated assistant that provides suggestion elements at an OEM application and/or third-party application that is accessible via an OEM vehicle computing device, which employs restrictions on communications between local applications. The automated assistant can render suggestions via OEM vehicle applications and/or third-party applications, and initialize actions for performance via the OEM vehicle applications and/or the third-party applications—despite the limitations. The suggestions can be rendered at a graphical user interface of a vehicle, and include content that, when spoken by a user, causes the automated assistant to initialize such actions. Actions that are suggested by the automated assistant can be performed by third-party applications that are different from a third-party application that is rendering the suggestion.
    Type: Grant
    Filed: May 9, 2019
    Date of Patent: May 4, 2021
    Assignee: GOOGLE LLC
    Inventors: Vinod Krishnan, Vikram Aggarwal
  • Patent number: 10997167
    Abstract: One of the methods includes, obtaining a natural language query from a user; while converting the natural language query into structured operations to be performed on structured application programming interfaces (APIs) of a knowledge base: responsive to detecting a linguistic ambiguity associated with the converting, wherein the linguistic ambiguity identifies two or more interpretations of a word included in the natural langue query: resolving the linguistic ambiguity by selecting an interpretation of the word, from the two or more interpretations of the word, based on one or more join paths that have been generated based on a data schema of the knowledge base; modifying the converting in accordance with the selected interpretation of the word, to generate one or more structured operations; performing the one or more structured operations on the structured APIs of the knowledge base; and returning search results matching the natural language query results to the user.
    Type: Grant
    Filed: March 14, 2019
    Date of Patent: May 4, 2021
    Assignee: Google LLC
    Inventors: Tolga Bozkaya, Armand Joseph Dijamco
  • Patent number: 10997368
    Abstract: A method includes generating a plurality of entigen groups from a set of phrases of a statement and identifying two plausible entigen groups based on a true meaning interpretation of the statement. The method further includes identifying a related entigen group based on a phrase of the statement and interpreting each of the two plausible entigen groups in light of the related entigen group to determine whether one of the two plausible entigen groups is a more likely interpretation of the statement than the other one of the two plausible entigen groups. When the one of the two plausible entigen groups is the more likely interpretation of the statement, the method further includes updating the one of the two plausible entigen groups using the related entigen group to produce an updated entigen group and adding the statement as the updated entigen group to a knowledge database.
    Type: Grant
    Filed: March 21, 2019
    Date of Patent: May 4, 2021
    Assignee: entigenlogic LLC
    Inventors: Frank John Williams, David Ralph Lazzara, Donald Joseph Wurzel, Stephen Chen, Dennis Arlen Roberson, Andrew Chu, Gary W. Grube