Machine Learning Patents (Class 706/12)
  • Patent number: 10296647
    Abstract: The present teaching relates to searching. In one example, a search query is received from a person. A plurality of search results are retrieved based on the search query. An intent of the person is estimated with respect to at least some of the plurality of search results. The estimated intent is what the person intends to do with respect to the at least some of the plurality of search results. The plurality of search results are provided based on the estimated intent of the person.
    Type: Grant
    Filed: October 5, 2015
    Date of Patent: May 21, 2019
    Assignee: OATH INC.
    Inventors: Jonathan Paris, Reiner Kraft
  • Patent number: 10296826
    Abstract: A system configured for learning new trained concepts used to retrieve content relevant to the concepts learned. The system may comprise one or more hardware processors configured by machine-readable instructions to obtain one or more digital media items. The one or more hardware processors may be further configured to obtain an indication conveying a concept to be learned from the one or more digital media items. The one or more hardware processors may be further configured to receive feedback associated with individual ones of the one or more digital media items. The one or more hardware processors may be configured to obtain individual neural network representations for the individual ones of the one or more digital media items. The one or more hardware processors may be configured to determine a trained concept based on the feedback and the neural network representations of the one or more digital media items.
    Type: Grant
    Filed: August 6, 2015
    Date of Patent: May 21, 2019
    Assignee: CLARIFAI, INC.
    Inventor: Matthew D. Zeiler
  • Patent number: 10296430
    Abstract: Mobile phones and methods for mobile phone failure prediction include receiving respective log files from one or more mobile phone components, including at least one user application. The log files have heterogeneous formats. A likelihood of failure of one or more mobile phone components is determined based on the received log files by clustering the plurality of log files according to structural log patterns and determining feature representations of the log files based on the log clusters. A user is alerted to a potential failure if the likelihood of component failure exceeds a first threshold. An automatic system control action is performed if the likelihood of component failure exceeds a second threshold.
    Type: Grant
    Filed: April 4, 2017
    Date of Patent: May 21, 2019
    Assignee: NEC Corporation
    Inventors: Jianwu Xu, Ke Zhang, Hui Zhang, Renqiang Min, Guofei Jiang
  • Patent number: 10296552
    Abstract: The system allows identifying new Internet advertising with minimal human participation. Also, the system provides a module for heuristic generation of rules for the found advertising, which makes it possible to automate the process of maintaining the relevance of lists of Internet advertising blocking rules. The system can operate in two modes: the mode of finding new advertising and generating rules for it and the mode of automatic data collection (datasets) for learning. The distinctive feature of the system is a comprehensive approach to the analysis of ad units that includes visual appearance of images (color contrast, element layout patterns, etc.), link analysis, and html code analysis.
    Type: Grant
    Filed: June 30, 2018
    Date of Patent: May 21, 2019
    Assignee: FiaLEAF LIMITED
    Inventors: Pavlo Malin, Oleksandr Chalyi, Oleksii Zinziuk, Volodymyr Shelest, Ivan Slieptsov
  • Patent number: 10296616
    Abstract: A processing device performs a preliminary grouping of data items in a dataset to define one or more clusters and for each cluster, identifies a set of search terms for a search query that would retrieve data items in the cluster upon execution of the search query against the dataset.
    Type: Grant
    Filed: July 31, 2014
    Date of Patent: May 21, 2019
    Assignee: SPLUNK INC.
    Inventors: Alice Neels, Steve Zhang, Marc Robichaud
  • Patent number: 10296833
    Abstract: Methods and arrangements for forecasting at least one attribute of a future event based on a repository of historical event data associated with historical events comparable to the future event. A plurality of event data points are obtained from the repository of historical event data. The plurality of event data points are grouped in accordance with at least one category and a plurality of subcategories to create at least one data tree. Certain of the grouped event data points are designated to form a set of candidate data attributes, and the designated set of candidate data attributes are compared to a set of data attributes from the at least one data tree associated with the future event. Based on the comparing, there is identified a data attribute missing from the set of data attributes from the at least one data tree associated with the future event, and a value for the missing data attribute is forecast. Other variants and embodiments are broadly contemplated herein.
    Type: Grant
    Filed: September 4, 2015
    Date of Patent: May 21, 2019
    Assignee: INTERNATIONAL BUSINESS MACHINES CORPORATION
    Inventors: Prithu Banerjee, Biplav Srivastava, Srikanth Govindaraj Tamilselvam
  • Patent number: 10296827
    Abstract: A deep neural network to which data category information is added is established locally, to-be-identified data is input to an input layer of the deep neural network generated based on the foregoing data category information, and information of a category to which the to-be-identified data belongs is acquired, where the information of the category is output by an output layer of the deep neural network. A deep neural network is established based on data category information, such that category information of to-be-identified data is conveniently and rapidly obtained using the deep neural network, thereby implementing a category identification function of the deep neural network, and facilitating discovery of an underlying law of the to-be-identified data according to the category information of the to-be-identified data.
    Type: Grant
    Filed: November 18, 2015
    Date of Patent: May 21, 2019
    Assignee: HUAWEI TECHNOLOGIES CO., LTD.
    Inventors: Guangjian Tian, Cheng He, Wei Fan
  • Patent number: 10298609
    Abstract: A method, system and computer-usable medium are disclosed for generating a cyber behavior profile comprising monitoring user interactions between a user and an information handling system; converting the user interactions into electronic information representing the user interactions, the electronic information representing the user interactions comprising multi-layered electronic information, each layer of the multi-layered electronic information corresponding to a respective layer of user interaction; and generating a unique multi-dimensional cyber behavior profile based upon the multi-layered electronic information representing the user interactions.
    Type: Grant
    Filed: May 8, 2018
    Date of Patent: May 21, 2019
    Assignee: Forcepoint, LLC
    Inventors: Richard Anthony Ford, Brandon L. Swafford
  • Patent number: 10289981
    Abstract: Methods, systems and computer program products are provided for computing relative score and enhancing one or more scores associated with data objects. In one method, the method receives, at a computing system, one or more data objects from a user. The method further extracts one or more internal and external parameters based on the one or more data objects. Subsequently, the method compares data objects of the user with corresponding data objects of one or more competing users, based at least in part on the one or more extracted parameters and the requirement included in the data object of the user. Further, in some embodiments, the data objects and requirements of the user are also compared with the corresponding data objects and requirements of one or more complementary users, based at least in part on the one or more extracted parameters.
    Type: Grant
    Filed: April 16, 2015
    Date of Patent: May 14, 2019
    Assignee: Microsoft Technology Licensing, LLC
    Inventors: Prabhakaran Krishnamoorthy, Uday Sankar Sen
  • Patent number: 10290039
    Abstract: A system and method is disclosed for profiling subjects and objects based on subjects' responses to various objects, for purposes of determining and presenting the objects most likely to generate the most likely response from each visitor. Object ratings are explicitly submitted by subjects or derived implicitly from visitor interactions with the objects. A profiling engine processes the ratings information and generates compact profiles of each subject and object based on similarities and differences in affinities between the group of subjects and the group of objects. The object profiles can be clustered to create behavioral object categories. Additionally, a modeling module inverts the abstract subject and object profiles into marketing attributes. The system has application in market analysis and segmentation, behavioral targeting, product placement, and online advertising, to name but a few applications.
    Type: Grant
    Filed: April 1, 2009
    Date of Patent: May 14, 2019
    Assignee: CERTONA CORPORATION
    Inventors: Geoffrey J. Hueter, Steven C. Quandt
  • Patent number: 10289819
    Abstract: Embodiments herein disclose a method and system for actively authenticating users of an electronic device in a continuous manner using a plurality of factors comprising of biometric modalities, power consumption, application usage, user interactions, user movement, and user location/travel.
    Type: Grant
    Filed: August 12, 2016
    Date of Patent: May 14, 2019
    Assignee: KRYPTOWIRE LLC
    Inventors: Angelos Stavrou, Rahul Murmuria, Ryan Johnson, Daniel Barbara
  • Patent number: 10289509
    Abstract: Methods for system failure prediction include clustering log files according to structural log patterns. Feature representations of the log files are determined based on the log clusters. A likelihood of a system failure is determined based on the feature representations using a neural network. An automatic system control action is performed if the likelihood of system failure exceeds a threshold.
    Type: Grant
    Filed: April 4, 2017
    Date of Patent: May 14, 2019
    Assignee: NEC Corporation
    Inventors: Jianwu Xu, Ke Zhang, Hui Zhang, Renqiang Min, Guofei Jiang
  • Patent number: 10290221
    Abstract: A computer implemented systems and methods for determining an action for a user within a learning domain are disclosed, some embodiments of the methods comprise defining an initial learning model of a learning domain, determining an initial user state of the user, determining an initial user action from at least one learning domain action with the initial learning model, receiving a user observation of the user after the user executes the initial user action, determining an updated user state with the initial learning model given the updated user observation and determining a subsequent user action from the at least one learning domain action. Some embodiments utilize a Partially Observable Markov Model (POMDP) as the learning model.
    Type: Grant
    Filed: April 29, 2013
    Date of Patent: May 14, 2019
    Assignee: Aptima, Inc.
    Inventors: E. Webb Stacy, Courtney Dean, Alan Carlin, Danielle Dumond
  • Patent number: 10289751
    Abstract: Discovery of causal networks is essential for understanding and manipulating complex systems in numerous data analysis application domains. Several methods have been proposed in the last two decades for solving this problem. The inventive method uses local causal discovery methods for global causal network learning in a divide-and-conquer fashion. The usefulness of the invention is demonstrated in data capturing characteristics of several domains. The inventive method outputs more accurate networks compared to other discovery approaches.
    Type: Grant
    Filed: March 17, 2014
    Date of Patent: May 14, 2019
    Inventors: Konstantinos (Constantin) F. Aliferis, Alexander Statnikov
  • Patent number: 10289909
    Abstract: A method and apparatus for classifying an image. In one example, the method may include receiving one or more images associated with a source domain and one or more images associated with a target domain, identifying one or more source domain features based on the one or more images associated with the source domain, identifying one or more target domain features based on the one or more images associated with the target domain, training a conditional maximum mean discrepancy (CMMD) engine based on a difference between the one or more source domain features and the one or more target domain features, applying the CMMD engine to the one or more images associated with the target domain to generate one or more labels for each unlabeled target image of the one or more images associated with the target domain and classifying each one of the one or more images in the target domain using the one or more labels.
    Type: Grant
    Filed: March 6, 2017
    Date of Patent: May 14, 2019
    Assignee: Xerox Corporation
    Inventors: Fabien Baradel, Boris Chidlovskii, Gabriela Csurka
  • Patent number: 10282462
    Abstract: A multi-modal computer classification network system for use in classifying data records is described herein. The system includes a memory device, a first classification computer server, a second classification computer server, and a policy computer server. The memory device includes an item records database and a labeling database. The first classification computer server includes a first classifier program that is configured to select an item record from the item database and generate a first classification record including a first ranked list of class labels. The second classification computer server includes a second classifier program that is configured to generate a second classification record including a second ranked list of class labels. The policy computer server includes a policy network that is programmed to determine a predicted class label based on the first and second ranked lists of class labels.
    Type: Grant
    Filed: October 31, 2016
    Date of Patent: May 7, 2019
    Assignee: WALMART APOLLO, LLC
    Inventors: Alessandro Magnani, Tom Ben Zion Zahavy, Abhinandan Krishnan, Shie Mannor
  • Patent number: 10282641
    Abstract: Technologies for classification using sparse coding are disclosed. A compute device may include a pattern-matching accelerator, which may be able to determine the distance between an input vector (such as an image) and several basis vectors of an overcomplete dictionary stored in the pattern-matching accelerator. The pattern matching accelerator may be able to determine each of the distances simultaneously and in a fixed amount of time (i.e., with no dependence on the number of basis vectors to which the input vector is being compared). The pattern-matching accelerator may be used to determine a set of sparse coding coefficients corresponding to a subset of the overcomplete basis vectors. The sparse coding coefficients can then be used to classify the input vector.
    Type: Grant
    Filed: July 1, 2016
    Date of Patent: May 7, 2019
    Assignee: Intel Corporation
    Inventors: Nilesh K. Jain, Soonam Lee
  • Patent number: 10282897
    Abstract: A method of automatically generating a three-dimensional entity is described. A sequence is generated comprising sets of blend shapes in order of increasing priority. Each set of blend shapes comprises one or more blend shape identifiers and parameters defining candidate blend weights for each blend shape. For each of the sets of blend shapes in the sequence and in order of increasing priority: the one or more blend shape identifiers in the set are added to a set of blend shapes for the entity. Dependent upon whether blend shape identifier that is added is already present in that set, the parameters for the blend shape are either added or updated. One or more blend shapes from the set of blend shapes for the entity are then added to a mesh of the entity using a randomization method and the resultant mesh is stored in memory.
    Type: Grant
    Filed: February 22, 2017
    Date of Patent: May 7, 2019
    Assignee: Microsoft Technology Licensing, LLC
    Inventors: Sam Chester, Daniel James Chalk, Iain McFadzen
  • Patent number: 10282546
    Abstract: The disclosed computer-implemented method for detecting malware based on event dependencies may include (1) applying, to a malware detection system capable of analyzing event dependencies, an event sequence derived from the execution of an application, (2) obtaining, from the malware detection system, a malware confidence score for the event sequence which the malware detection system calculates after a certain event within the event sequence has executed based at least in part on one or more dependencies between the certain event and at least one other event within the event sequence, (3) determining that the malware confidence score exceeds a threshold, and (4) classifying the application as malicious in response to determining that the malware confidence score exceeds the threshold. Various other methods, systems, and computer-readable media are also disclosed.
    Type: Grant
    Filed: June 21, 2016
    Date of Patent: May 7, 2019
    Assignee: Symatec Corporation
    Inventors: Jugal Parikh, Reuben Feinman
  • Patent number: 10282443
    Abstract: In an approach for query processing, a computer receives a query. The computer determines the received query does not correspond to a previously executed query. The computer parses the received query to identify input literals that include one or more of: data values, tables, fields, records, and a parameter included in the received query. The computer determines whether a pattern is associated with the input literals included in the received query. Responsive to determining a pattern is associated with the input literals, the computer determines a future parameter based on the pattern associated with the input literals, wherein the future parameter is a subsequent instance of a parameter of the received query that corresponds to the pattern. The computer creates a future query by incorporating the determined future parameter into the received query.
    Type: Grant
    Filed: March 2, 2016
    Date of Patent: May 7, 2019
    Assignee: International Business Machines Corporation
    Inventors: Ge Song, Kewei Wei, Xin Ying Yang
  • Patent number: 10282237
    Abstract: Systems and methods for implementing an application programming interface (API) that controls operations of a machine learning tuning service for tuning a machine learning model for improved accuracy and computational performance includes an API that is in control communication the tuning service that: executes a first API call function that includes an optimization work request that sets tuning parameters for tuning hyperparameters of a machine learning model; and initializes an operation of distinct tuning worker instances of the service that each execute distinct tuning tasks for tuning the hyperparameters; executes a second API call function that identifies raw values for the hyperparameters; and generates suggestions comprising proposed hyperparameter values selected from the plurality of raw values for each of the hyperparameters; and executes a third API call function that returns performance metrics relating to a real-world performance of the subscriber machine learning model executed with the propose
    Type: Grant
    Filed: October 29, 2018
    Date of Patent: May 7, 2019
    Assignee: SigOpt, Inc.
    Inventors: Alexandra Johnson, Patrick Hayes, Scott Clark
  • Patent number: 10282668
    Abstract: A system and method of confirming compliance with rules. A machine learning model is trained in multiple levels with data concerning basic interpretation of sensors and with discovery of interesting patterns related to the area of application of the rules. The model is downloaded to a second processor which further trains the model with sensor data gathered after the download. Additional sensor data along with data from a server or other sources may be used as an input to the model and the second or another processor evaluates the model with the inputs to create outputs which determine the state of compliance. Specific applications include determination of compliance with a dress code and determination of sobriety in traffic stop situations. Other typical applications include compliance with sports regulations such as the one requiring a specific geometry of a lacrosse stick head pocket as checked after each goal.
    Type: Grant
    Filed: January 31, 2018
    Date of Patent: May 7, 2019
    Inventor: Thomas Danaher Harvey
  • Patent number: 10275628
    Abstract: A data-analytics application may be optimized for implementation on a computing device for conserving computing or providing timely results, such as a prediction, recommendation, inference, or diagnosis about a monitored system, process, event, or a user, for example. A feature filter or classifier is generated and incorporated into or used by the application to provide the optimization. The feature filter and classifier are generated based on a set of significant features, determined using a data condensation and summarization process, from a high-dimensional set of available features characterizing the target. For example, a process that includes utilizing combined sparse principal component analysis with sparse singular value decomposition and applying k-medoids clustering may determine the significant features. Insignificant features may be filtered out or not used, as information represented by the insignificant features is expressed by the significant features.
    Type: Grant
    Filed: May 27, 2016
    Date of Patent: April 30, 2019
    Assignee: Adobe Inc.
    Inventor: Kourosh Modarresi
  • Patent number: 10275473
    Abstract: A generative adversarial networks-based or GAN-based method for learning cross-domain relations is disclosed. A provided architecture includes two coupled GANs: a first GAN learning a translation of images from domain A to domain B, and a second GAN learning a translation of images from domain B to domain A. A loop formed by the first GAN and the second GAN causes sample images to be reconstructed into an original domain after being translated into a target domain. Therefore, loss functions representing reconstruction losses of the images may be used to train generative models.
    Type: Grant
    Filed: August 9, 2017
    Date of Patent: April 30, 2019
    Assignee: SK TELECOM CO., LTD.
    Inventors: Taek Soo Kim, Moon Su Cha, Ji Won Kim
  • Patent number: 10275719
    Abstract: Hyper-parameters are selected for training a deep convolutional network by selecting a number of network architectures as part of a database. Each of the network architectures includes one or more local logistic regression layer and is trained to generate a corresponding validation error that is stored in the database. A threshold error for identifying a good set of network architectures and a bad set of network architectures may be estimated based on validation errors in the database. The method also includes choosing a next potential hyper-parameter, corresponding to a next network architecture, based on a metric that is a function of the good set of network architectures. The method further includes selecting a network architecture, from among next network architectures, with a lowest validation error.
    Type: Grant
    Filed: September 8, 2015
    Date of Patent: April 30, 2019
    Assignee: QUALCOMM Incorporated
    Inventors: Sachin Subhash Talathi, David Jonathan Julian
  • Patent number: 10275513
    Abstract: A system is configured to obtain data for a user. The data may describe actions that the user has performed in an application and identify points in time associated with the actions. A point in time may include a point in time at which one of the actions was performed. The system is configured to analyze the data to determine, for each action, a score for each point in time. The score, for a particular point in time, may be determined based on one or more points in time at which the action was performed, and may be determined from a number of times that the action was performed at the particular point in time. The system is configured to provide the user with a functionality to perform a particular action based on the scores determined for the particular action for the set of points in time.
    Type: Grant
    Filed: June 27, 2013
    Date of Patent: April 30, 2019
    Assignee: GOOGLE LLC
    Inventors: Mathew Cowan, Christopher Pedregal
  • Patent number: 10275481
    Abstract: In updating a synopsis table of a database system, a database management unit performs a transaction to insert row(s) in a section of the base table and determines whether a synopsis entry for the section is stored in the memory. If stored in the memory, the in-memory synopsis entry is retrieved and metadata values in the in-memory synopsis entry are updated with data from the row(s) to be inserted. If not stored in the memory, the in-memory synopsis entry is generated and the metadata values in the in-memory synopsis entry are updated with data from the row(s). The insert transaction is then committed. Synopsis entry on-disk updates are thus avoided, significantly reducing the cost of updating the synopsis entries from the insert transaction. This yields enhanced performance especially for inserts of a small number of rows, while the benefits of synopsis entries are still available.
    Type: Grant
    Filed: October 11, 2016
    Date of Patent: April 30, 2019
    Assignee: International Business Machines Corporation
    Inventors: James L. Finnie, Sam S Lightstone, Richard S Sidle, Adam J. Storm
  • Patent number: 10275301
    Abstract: An approach is provided for detecting and analyzing an anomaly in application performance in a client-server connection via a network. A request time and an Internet Protocol (IP) address of the client are determined. Based on the request time and the IP address, log entries relevant to the request are selected. A response code, a round trip latency time (RTT) of the response, and an indication of whether the connection timed out are determined. Based on the status code, the RTT, and the indication of whether connection timed out, the anomaly is detected. Based on temporal and textual analyzes of log entries associated with the anomaly and an environment analysis that determines activity of the client, server, and network, candidate root causes of a failure that resulted in the anomaly are determined.
    Type: Grant
    Filed: September 29, 2015
    Date of Patent: April 30, 2019
    Assignee: International Business Machines Corporation
    Inventors: Luba Cherbakov, Kuntal Dey, Sougata Mukherjea, Nitendra Rajput, Venkatraman Ramakrishna
  • Patent number: 10275418
    Abstract: Selecting and ranking valid variants in search and recommendation systems selects and ranks variants with accuracy and speed. Criteria for evaluating the relevance of a variant to the search request are generated. A set of procedures for the selection and ranking of variants and a sequence for performing said procedures for the selection of variants evaluated as the most valid are established. An evaluation of each variant is based on relevance to search request criteria. The variants are then ranked by assigning a rank to each variant based on the condition of correspondence to the greatest number of criteria in decreasing order. Then the variants are selected and ranked in at least two stages using the superposition method, and the variants are selected, ranked and excluded until all of the established selection procedures have been used and the selected group of variants is evaluated as being the most valid.
    Type: Grant
    Filed: September 17, 2015
    Date of Patent: April 30, 2019
    Assignee: National Research University Higher School of Economics (HSE)
    Inventors: Fuad T. Aleskerov, Evgeny O. Mitichkin, Vyacheslav V. Chistyakov, Sergey V. Shvydun, Viacheslav I. Iakuba
  • Patent number: 10268461
    Abstract: A method for global data flow optimization for machine learning (ML) programs. The method includes receiving, by a storage device, an initial plan for an ML program. A processor builds a nested global data flow graph representation using the initial plan. Operator directed acyclic graphs (DAGs) are connected using crossblock operators according to inter-block data dependencies. The initial plan for the ML program is re-written resulting in an optimized plan for the ML program with respect to its global data flow properties. The re-writing includes re-writes of: configuration dataflow properties, operator selection and structural changes.
    Type: Grant
    Filed: November 23, 2015
    Date of Patent: April 23, 2019
    Assignee: International Business Machines Corporation
    Inventors: Matthias Boehm, Mathias Peters, Berthold Reinwald, Shirish Tatikonda
  • Patent number: 10268969
    Abstract: A process and computer program product to record performance related data for a plurality of entertainment performances having a plurality of audiences. Further, the process and computer program product determine external data that is associated with the plurality of audiences and/or environmental factors corresponding to locations of the plurality of entertainment performances. In addition, the process and computer program product annotate the performance related data with the external data to generate annotated performance related data. The process and computer program product also train an artificial intelligence system based upon the annotated performance related data. The process and computer program product generate, at the artificial intelligence performance instructions to provide a performance. Further, the process and computer program product provide, from the artificial intelligence system to a performance device, the performance instructions to provide a performance.
    Type: Grant
    Filed: April 16, 2015
    Date of Patent: April 23, 2019
    Assignee: Disney Enterprises, Inc.
    Inventors: Jon Snoddy, Alfredo Ayala, Daniel Pike, Mike Fusco, Douglas Fidaleo
  • Patent number: 10265040
    Abstract: The invention provides a method and apparatus for classifying a region of interest in imaging data, the method comprising: calculating a feature vector for at least one region of interest in the imaging data; projecting the feature vector for the at least one region of interest in the imaging data using a plurality of decision functions to generate a corresponding plurality of classifications; calculating an ensemble classification based on the plurality of classifications. receiving from the user feedback information concerning the ensemble classification; forming an additional classified feature vector from the feature vector and the feedback information; and updating at least one of the plurality of decision functions using the additional classified feature vector.
    Type: Grant
    Filed: July 13, 2015
    Date of Patent: April 23, 2019
    Inventors: Scott Anderson Middlebrooks, Henricus Wilhelm van der Heijden
  • Patent number: 10271020
    Abstract: Systems can include a tool configured to generate a first set of data representative of an object and a control station configured to receive the first set of data from the tool. A mobile device in communication with the control station can receive data from the control station including information regarding the object based on the data received from the tool. A user of the mobile device, such as a technician, can travel to the location of the tool, and, when the mobile device is within a predetermined proximity of the tool, the tool can communicate directly with the mobile device. The technician can use the mobile device to communicate with and control the tool in order to safely perform equipment analysis.
    Type: Grant
    Filed: October 23, 2015
    Date of Patent: April 23, 2019
    Assignee: Fluke Corporation
    Inventors: Michael D. Stuart, Michael A. Schoch
  • Patent number: 10262233
    Abstract: The presence of possibility of occurrence of an excessive adaptation due to use of only learned training data is detected during a learning stage. The user is urged to add data and other information, thereby avoiding the excessive adaptation. For this purpose, the invention has: an inputting unit for inputting a learning image; a generating unit for generating a discrimination model used to decide whether or not a target is normal on the basis of the learning image; a deciding unit for deciding whether or not the number of input learning images is insufficient when the discrimination model is generated; and a notifying unit for notifying a message for urging the user to add the learning image when it is decided that the number of input learning images is insufficient.
    Type: Grant
    Filed: January 12, 2016
    Date of Patent: April 16, 2019
    Assignee: CANON KABUSHIKI KAISHA
    Inventor: Masafumi Takimoto
  • Patent number: 10261959
    Abstract: Systems and methods for converting a data item provided by an external data provider system into a data type specified by a data processing system for a data field of the data item. A data processing system stores a coercion rule for each data field of a first data set provided by the data provider system. Each stored coercion rule identifies at least one data type for the corresponding data field. Responsive to a second data set provided by the data provider system, the data processing system coerces each data item of the second data set into at least one data type specified by the stored coercion rule for the data field of the data item to generate at least one converted data item of the second data set. The data processing system generates information from at least one converted data item, and provides the information to a consuming system.
    Type: Grant
    Filed: June 28, 2017
    Date of Patent: April 16, 2019
    Assignee: ZestFinance, Inc.
    Inventors: John W. L. Merrill, John J. Beahan
  • Patent number: 10264048
    Abstract: In an example embodiment, a supervised machine learning algorithm is used to train a communication reply score model based on an extracted first set of features and second set of features from social networking service member profiles and activity and usage information. When a plurality of member search results is to be displayed, for the member identified in each of the plurality of member search results, the member profile corresponding to the member is parsed to extract a third set of one or more features from the member profile, activity and usage information pertaining to actions taken by the members on the social networking service is parsed to extract a fourth set of one or more features, and the extracted third set of features and fourth set of features is inputted into the communication reply score model to generate a communication reply score, which is displayed visually to a searcher.
    Type: Grant
    Filed: February 23, 2016
    Date of Patent: April 16, 2019
    Assignee: Microsoft Technology Licensing, LLC
    Inventors: Qiang Zhu, Qingbo Hu
  • Patent number: 10255628
    Abstract: A deep collaborative filtering (DCF) approach is employed in a recommender system to provide item recommendations to users. The DCF approach combines deep learning models with matrix factorization based collaborative filtering. To provide item recommendations, a user-item rating matrix, user side information, and item side information are provided as input to a recommender system. The recommender system learns user latent factors and item latent factors by jointly: (1) decomposing the user-item rating matrix to extract latent factors, and (2) extracting latent factors from hidden layers of deep learning models using the user side information and item side information. The learned user latent factors and item latent factors are used to predict item ratings for missing ratings in the user-item rating matrix. The predicted item ratings are then used to select item recommendations for a given user, which are then communicated to a user device of the user.
    Type: Grant
    Filed: November 6, 2015
    Date of Patent: April 9, 2019
    Assignee: Adobe Inc.
    Inventors: Sheng Li, Jaya Kawale
  • Patent number: 10254935
    Abstract: Systems and methods of providing content selection are provided. For instance, one or more signals indicative of a user selection of an object displayed within a user interface can be received. Responsive to receiving the one or more signals, a content attribute associated with one or more objects displayed within the user interface can be identified. A content entity can be determined based at least in part on the content attribute and the user selection. One or more relevant actions can then be determined based at least in part on the determined content entity. Data indicative of the relevant actions can then be provided for display.
    Type: Grant
    Filed: June 29, 2016
    Date of Patent: April 9, 2019
    Assignee: Google LLC
    Inventors: Stefano Mazzocchi, Kaikai Wang, John Thomas DiMartile, III, Tim Wantland
  • Patent number: 10255352
    Abstract: Described is a system for early detection of events via social media mining. The system receives, as input, social media blog posts comprising textual data. The system processes the social media blog posts through a cascade of filters. The cascade of filters comprises an event term detection filter, a location term detection filter following the event term detection filter, and a future date detection filter following the location term detection filter. A plurality of candidate social media blog posts describing an event of interest on a future date is output to a user for further analysis.
    Type: Grant
    Filed: November 7, 2014
    Date of Patent: April 9, 2019
    Assignee: HRL Laboratories, LLC
    Inventors: Jiejun Xu, Tsai-Ching Lu, Ryan F. Compton, David L. Allen
  • Patent number: 10255238
    Abstract: The present technology concerns a complex event processing (CEP) engine for processing CEP queries over data streams. The CEP engine has a parser, adapted for parsing a received CEP query into a logical query graph and a translator adapted for translating the logical query graph into a physical query plan in accordance with one of a plurality of data stream representations. The logical query graph is independent of the plurality of data stream representations.
    Type: Grant
    Filed: January 31, 2011
    Date of Patent: April 9, 2019
    Assignee: SOFTWARE AG
    Inventors: Michael Cammert, Christoph Heinz, Jürgen Krämer, Tobias Riemenschneider
  • Patent number: 10255807
    Abstract: An approach is provided for map data updates based on region-specific data turbulence. The approach involves, for example, retrieving historical map data for the map region and segmenting the historical map data into a time series including at least a first time epoch and a second time epoch. The approach also involves calculating a first representative value for the first time epoch based on the historical map data segmented into the first time epoch, and a second representative value for the second time epoch based on the historical map data segmented into the second time epoch. The approach further involves calculating the map data turbulence based on the first representative value and the second representative value.
    Type: Grant
    Filed: September 28, 2017
    Date of Patent: April 9, 2019
    Assignee: HERE Global B.V.
    Inventor: Leon Stenneth
  • Patent number: 10255345
    Abstract: A raw dataset including measures and dimensions is processed, by a preprocessing module, using an algorithm that produces a preprocessed dataset such that at least one type of statistical analysis of the preprocessed dataset yields equal results to the same type of statistical analysis of the raw dataset. The preprocessed dataset is then analyzed by a statistical analysis module to identify subsets of the preprocessed dataset that include a non-random structure or pattern. The analysis of the preprocessed dataset includes the at least one type of statistical analysis that produces the same results for both the preprocessed and raw datasets. The identified subsets are then ranked by a statistical ranker based on the analysis of the preprocessed dataset and a subset is selected for visualization based on the rankings. A visualization module then generates a visualization of the selected identified subset that highlights a non-random structure of the selected subset.
    Type: Grant
    Filed: October 9, 2014
    Date of Patent: April 9, 2019
    Assignee: Business Objects Software Ltd.
    Inventors: Flavia Moser, Alexander Kennedy MacAulay, Julian Gosper
  • Patent number: 10257572
    Abstract: Introduced herein are methods and systems for determining machine learning marketing strategy. For example, a computer-implemented method according to the disclosed technology includes steps of identifying one or more business metrics to be driven by a marketing plan; generating one or more response functions of the business metrics by performing a machine learning process on a marketing dataset; optimizing a spending subject of the marking plan subject to constraints to generate a marketing strategy based on multiple decision variables; and presenting the marketing strategy to an advertiser.
    Type: Grant
    Filed: April 24, 2017
    Date of Patent: April 9, 2019
    Assignee: Bliss Point Media, Inc.
    Inventors: Justin Manus, Sean Odlum, Anand V. Bodapati
  • Patent number: 10257306
    Abstract: Embodiments of the present invention provide a service data cache processing method and system and a device. The method includes receiving statistical information of service data and sending a service data push request to a service provider SP device according to the statistical information, so that the SP device sends service data to a primary cache deployed in a core network or an edge cache deployed in an access network.
    Type: Grant
    Filed: March 12, 2015
    Date of Patent: April 9, 2019
    Assignee: Huawei Technologies Co., Ltd.
    Inventors: Anni Wei, Chunshan Xiong
  • Patent number: 10248667
    Abstract: A real-time messaging platform allows advertiser accounts to pay to insert candidate messages into the message streams requested by account holders. To accommodate multiple advertisers, the messaging platform controls an auction process that determines which candidate messages are selected for inclusion in a requested account holder's message stream. Selection is based on a bid for the candidate message, the message stream that is requested, and a variety of other factors that vary depending upon the implementation. The process for selection of candidate messages generally includes the following steps, though any given step may be omitted or combined into another step in a different implementation: targeting, filtering, prediction, ranking, and selection.
    Type: Grant
    Filed: March 14, 2014
    Date of Patent: April 2, 2019
    Assignee: TWITTER, INC.
    Inventors: Parag Agrawal, Mike Jahr, Yue Lu, Ke Zhou, Utkarsh Srivastava
  • Patent number: 10242320
    Abstract: A data model is traversed to determine concept characteristics associated with concepts that may be associated with entities. Associated documents may be evaluated to identify document characteristics associated with the entities. Entity models may be trained based on the concept characteristics and the document characteristics with each entity model being associated with a confidence value. Results for one or more queries based on the documents and the entity models may be provided. The results may reference the documents that may be associated with the entities. Some entity models may produce results that have a confidence value below a threshold value. Accordingly, the entity models that provide low confidence results may be re-trained.
    Type: Grant
    Filed: April 19, 2018
    Date of Patent: March 26, 2019
    Assignee: Maana, Inc.
    Inventors: Alexander Hussam Elkholy, Balasubramanian Kandaswamy, Steven Matt Gustafson, Hussein S. Al-Olimat
  • Patent number: 10244286
    Abstract: Methods and apparatuses are described for recommending digital content in a network environment. A web server determines a session context of a user associated with a remote device in which the user interacted with digital content objects. The web server identifies prior content interactions in which prior users interacted with digital content objects. The web server builds a recommendation feature vector from the session context and the prior content interactions. The web server trains a machine learning model using the generated recommendation feature vector to minimize a predicted rating. The web server selects target digital content objects that have a predicted rating of the user that is at a threshold. The web server provides indicia of the selected digital content objects to the remote device. The web server causes at least one of the selected target digital content objects to be transmitted from a content server to the remote device.
    Type: Grant
    Filed: January 30, 2018
    Date of Patent: March 26, 2019
    Assignee: FMR LLC
    Inventor: Yugang Jia
  • Patent number: 10242207
    Abstract: For preventing unwanted information disclosure in a current electronic communication from a sender to a receiver of the current electronic communication, a risk score is assigned for the current electronic communication by a computer system applying an access control model based on historical electronic communications. The model generates the risk score responsive to identities of the sender and receiver and responsive to access control level and hierarchal position of at least one of the sender and receiver. The computer system blocks transmission of the current electronic communication from the sender to the receiver responsive to whether the risk score for the sender and receiver exceeds a predetermined threshold.
    Type: Grant
    Filed: September 17, 2016
    Date of Patent: March 26, 2019
    Assignee: INTERNATIONAL BUSINESS MACHINES CORPORATION
    Inventors: Hernan A. Cunico, Jonathan Dunne, Jeremiah O'Connor, Asima Silva
  • Patent number: 10242114
    Abstract: A method is provided of enriching an entry for an entity in a local index of a search engine with tags. The method comprises obtaining location-related social media messages from within a neighborhood of an entity; determining from the obtained messages one or more terms that are unique to the entity; individually determining one or more co-occurring terms for the one or more unique terms; and using the one or more co-occurring term as tags to label the entity in the local index. Furthermore, a method is provided of retrieving social media messages associated with search results.
    Type: Grant
    Filed: December 30, 2013
    Date of Patent: March 26, 2019
    Assignee: Microsoft Technology Licensing, LLC
    Inventors: Riham Hassan Abdel-Moneim Mansour, Joseph W. Pepper, Nesma Abd El-Hakim Refaei, Diaa Mohamed Abdel Moneim Abdallah, Vanessa Graham Murdock
  • Patent number: 10242264
    Abstract: A method and a system for training a machine-learning model to identify real-world elements using a simulated environment (SE) may include (a) receiving at least one set of appearance parameters, corresponding to appearance of real-world element; (b) generating one or more realistic elements, each corresponding to a variant of at least one real-world element; (c) generating one or more abstract-elements; (d) placing the elements within the SE; (e) producing at least one synthetic image from the SE; (f) providing the at least one synthetic image to a machine-learning model; and (g) training the machine-learning model to identify at least one real-world element from the at least one synthetic image, that corresponds to at least one realistic element in the SE.
    Type: Grant
    Filed: April 25, 2018
    Date of Patent: March 26, 2019
    Assignee: IMAGRY (ISRAEL) LTD.
    Inventors: Sergey Ten, Jose Ariel Keselman, Suhail Habib, Abed Abu Dbai, Adham Ghazali, Majed Jubeh, Itai Orr