International Business Machines Corporation

Sumit Patel

Liam S Harpur

Erik H Katzen

Al Chakra

David Bacarella

John Rice

Romelia H Flores

Nicholas Todd Lawrence

James H Barnebee, Iv

Thomas Edward Creamer

Christina I Flores

Jonathan D Dunne

Enda Sexton

Julia Ruth Nash

Ross Warren Judd

Osakpamwan Osaigbovo

Pascal Wagner

Jeremy R Fox

Leigh Allen Williamson

Howard Neil Anglin

Stephan Joseph Roorda

Christopher M Barton

Arunava Majumdar

Gregor Schikorski

Ginés Carrascal De Las Heras

Christopher M Crane

Paulo Henrique Leal

Chandra Shekhar

Pattern recognition by receiving a set multi-variable data records, each record including a plurality of variables, representing at least two of the plurality of variables as geometric shapes, defining a boundary enclosing the geometric shapes, configuring at least one geometric shape to move within the boundary, capturing a location of each of the geometric shapes within the boundary as a system state, one or more times, combining one or more system states as a system signature, providing a model trained to recognize patterns in system signatures, and recognizing a pattern in the system signature.

Multi-variable pattern recognition for predictive deep learning models

An example operation may include one or more of authorizing a blockchain for a video file, generating a first tracking value for an entry block referencing the video file, the first tracking value generated based on first data and the video file, receiving second data for each of additional blocks in the blockchain, generating second tracking values based on the second data of the additional blocks, forming the additional blocks including the second tracking values, respectively, appending the additional blocks to the entry block, the entry block and the additional blocks cryptographically linked in an ordered sequence, each of the additional blocks referencing a version of the video file which corresponds to an original version of the video file as referenced by the entry block or a processed version of the video file, the second data in each of the additional blocks indicative of processing performed on the version of the video file in that block, and tracing through the blockchain based on the first and second tracking values to confirm an auditable and immutable chain-of-custody of the video file.

Chain-of-custody of digital content in a database system

Embodiments of the present invention provide a computer system, a computer program product, and a method that comprises determining user activity based on a personalized electronic history; generating a user interface that overlays on a physical object that denotes user interaction on the physical object based, at least in part, on the determined user activity and user location; and displaying the generated user interface that overlays on the physical object on the user interface on a computing device.

Automatic electronic history centralization

In an approach for providing an enhancement to message check capabilities through an awareness of the message content, a processor processes a message that includes a non-textual part and a textual part. A processor identifies a content context of the non-textual part of the message. A processor identifies an intent and entity of the textual part of the message. A processor determines a potential mismatch between the content context of the non-textual part and the intent and entity of the textual part. A processor provides a suggested replacement for the potential mismatch.

Content context aware message intent checker

A computer-implemented method includes: receiving, by a computing device, data corresponding to a dynamic speckle spectrum image associated with a biological sample; comparing, by the computing device, the dynamic speckle spectrum image with a plurality of training images; classifying, by the computing device, a contaminant present in the biological sample, based on the comparing; and providing, by the computing device, information regarding the classification of the contaminant.

Testing a biological sample based on sample spectrography and machine learning techniques

A presenter and an audience associated with a presentation event can be detected. The presentation event can be a real world occurrence and/or a computing session occurrence. The presenter can convey a presentation content to the audience. Presentation data associated with the presentation event can be identified. The presentation data can include the presentation content. The presentation content can include a text and/or a media. A real-time metrics associated with the presenter and the audience can be captured from one or more computing devices associated with the presenter and the audience during the presentation event. The real-time metrics can be analyzed to determine a presenter and an audience recommendation. The presenter recommendation can be presented to the presenter and the audience recommendation to an audience member.

Enhancing presentation content delivery associated with a presentation event

Embodiments for adding content to images are provided. A plurality of images are received. Each of the plurality of images includes a plurality of image portions. A negative space score is calculated for each of the plurality of image portions. At least some of the plurality of image portions are selected to display content based on the calculated negative space scores. The plurality of images are caused to be rendered with the content displayed over the selected at least some of the plurality of image portions.

Methods and systems for adding content to images based on negative space recognition

An automated dynamic message categorization system is provided and includes first, second and third processing units. The first processing unit is configured to generate a user interface (UI) and to present the UI to a user. The second processing unit is configured to pull information from a first textual element which has been entered into the UI, to identify second textual elements that are relevant to the first textual element based on the pulled information and to extract textual element identifiers from the second textual elements. The third processing unit is configured to generate, for each extracted textual element identifier, a confidence score describing a degree of correlation between each extracted textual element identifier and the first textual element. The first processing unit is further configured to present to the user each extracted textual element identifier with a corresponding confidence score as a selectable option via the UI.

Dynamic message categorization for optimized message targeting

A system, computer program product, and method are provided to automate a framework for knowledge graph based persistence of data, and to resolve temporal changes and uncertainties in the knowledge graph. Natural language understanding, together with one or more machine learning models (MLMs), is used to extract data from unstructured information, including entities and entity relationships. The extracted data is populated into a knowledge graph. As the KG is subject to change, the KG is used to create new and retrain existing machine learning models (MLMs). Weighting is applied to the populated data in the form of veracity value. Blockchain technology is applied to the populated data to ensure reliability of the data and to provide auditability to assess changes to the data.

Machine learning model modification and natural language processing

A computer-implemented method includes associating a tag with a message in a log, where the message describes a logged event. A life parameter associated with the tag is calculated, using a computer processor, where the life parameter indicates a dynamic time period during which the tag remains in a state that is active. Alerting of the tag occurs by way of a first alert manner while the state of the tag is active. It is detected that the dynamic time period indicated by the life parameter has ended. The state of the tag is changed from active to dormant, where changing the state includes ceasing the alerting of the tag by way of the first alert manner.

Prioritizing log tags and alerts

Machine learning to integrate knowledge and augment natural language processing

Monitoring oral intake of consumable items may include aggregating, using a processor, oral intake information for a user from a plurality of devices belonging to a same device domain of the user in real time and comparing, using the processor, the oral intake information with a user profile of the user. An instruction for a selected device of the plurality of devices may be generated, using the processor, according to the comparing. The instruction may be sent to the selected device, using the processor. The selected device may execute the instruction.

Monitoring and status detection for consumable items

Utilizing problem insights based on the entire environment as inputs to drive a static compiler. A decision engine receives inputs associated with applications to be compiled. The decision engine also receives optimization constraints based on available resources. A decision learning model is applied to the inputs to predict compiler performance and the results are provided to the decision engine. The decision engine determines a profile that comprises an order of execution and an optimization level for use during compilation of the plurality of applications. The profile is then used to schedule compiling and optimization of the applications.

Analytics driven compiler

In an approach for determining a preferred communication channel, the computer receives a query that includes a unique client identifier. The computer retrieves one or more communication channels from an entry within a data repository, wherein the entry is associated with the received query. The computer determines one or more communication channel results based, at least in part, on one or more success rates respectively associated with the retrieved one or more communication channels within the data repository. The computer creates a first ordered list of communication channel results in which the determined one or more communication channel results are ordered based, at least in part, on the one or more success rates of the determined one or more communication channels. The computer selects a preferred communication channel from the created first ordered list of communication channel results. The computer initiates a communication via the selected preferred communication channel.

Growing community of inventors

Irving, TX, United States of America

Sumit Patel