International Business Machines Corporation

Sas Institute Inc.

David B Styles

Elizabeth A Schreiber

Eric Woods

Scott Robert Carrier

Scott N Gerard

Sterling Richardson Smith

Corville O Allen

Barnaby L Court

Radhika Aggarwal

Brian John Cragun

William H Krebs, Jr

Michael Peter Etgen

Timothy Jude O'Keefe

Brian Owen Wood

Samuel Paul Leeman-Munk

Richard Welland Crowell

Andrew L Hanson

Radhika Bhatnagar

Anjan Reddy Kundavaram

James Allen Cox

Roland Albert Merrick

Reza Soleimani

David Bruce Lection

Clifton Malcolm Nock

Saratendu Sethi

Charles Edward Burkett

Jared Michael Dean Smythe

James Tetterton

Thomas Jay Allen

William H Krebs Jr

Viswanath Avasarala

In one example, a system can receive a set of text samples and generate a set of summaries based on the set of text samples. The system can then generate a training dataset by iteratively executing a training-sample generation process. Each iteration can involve selecting multiple text samples from the set of text samples, combining the multiple text samples together into a training sample, determining a text category and a summary corresponding to a selected one of the multiple text samples, and including the text category and the summary in the training sample. After generating the training dataset, the system can use it to train a model. The trained model can then receive a target textual dataset and a target category as input, identify a portion of the target textual dataset corresponding to the target category, and generate a summarization of the portion of that target textual dataset.

Machine-learning model for performing contextual summarization of text data

In one example, a system can execute a first machine-learning model to determine an overall classification for a textual dataset. The system can also determine classification scores indicating the level of influence that each token in the textual dataset had on the overall classification. The system can select a first subset of the tokens based on their classification scores. The system can also execute a second machine-learning model to determine probabilities that the textual dataset falls into various categories. The system can determine category scores indicating the level of influence that each token had on a most-likely category determination. The system can select a second subset of the tokens based on their category scores. The system can then generate a first visualization depicting the first subset of tokens color-coded to indicate their classification scores and a second visualization depicting the second subset of tokens color-coded to indicate their category scores.

Graphical user interface for visualizing contributing factors to a machine-learning model's output

A system trains a classification model. Text windows are defined from tokens based on a window size. A network model including a transformer network is trained with the text windows to define classification information. A first accuracy value is computed. (A) The window size is reduced using a predefined reduction factor value. (B) Second text windows are defined based on the reduced window size. (C) Retrain the network model with the second text windows to define classification information. (D) A second accuracy value is computed. (E) An accuracy reduction value is computed from the second accuracy value relative to the first accuracy value. When the computed accuracy reduction value is &#x2265;an accuracy reduction tolerance value, repeat (A)-(E) until the accuracy reduction value is &#x3c;the accuracy reduction tolerance value. Otherwise, increase the window size, define final text windows based on the increased window size, and retrain the network model.

Machine learning classification system

A system determines user intent from text. A conversation element is received. An intent is determined by matching a domain independent relationship and a domain dependent term determined from the received conversation element to an intent included in an intent database that stores a plurality of intents and by inputting the matched intent into a trained classifier that computes a likelihood that the matched intent is the intent of the received conversation element. An action is determined based on the determined intent. A response to the received conversation element is generated based on the determined action and output.

System for determining user intent from text

A search engine is enhanced to employ user preferences to customize answer output by, receiving a first plurality of candidate answers and weights to a first search query for a first user from a first stage of a non-semantic search engine routing and processing path; augmenting the weights for the plurality of candidate answers according to a set of preference coefficients according to a first set of preferences for the first user; selecting one or more of the first plurality of candidate answers according to the augmented weights; and producing an output including the selected one or more candidate answers.

Enhanced answer in DeepQA system according to user preferences

A semantic search engine is enhanced to employ user preferences to customize answer output by, for a first user, extracting user preferences and sentiment levels associated with a first question; receiving candidate answer results of a semantic search of the first question; weighting the candidate answer results according to the sentiment levels for each of the user preferences; and producing the selected candidate answers to the first user. Optionally, user preferences and sentiment levels may be accumulated over different questions for the same user, or over different users for similar questions. And, supplemental information may be retrieved relative to a user preference in order to further tune the weighting per the preferences and sentiment levels.

Enhanced answers in DeepQA system according to user preferences

Visualizable subject matter contained within data is detected. A context in which the visualizable subject matter appears is determined. A visualization engine of a plurality of visualization engines is selected based, at least in part, on the context in which the visualizable subject matter appears. The visualization engine is configured to generate a visualization of the visualizable subject matter.

Context-based visualization generation

In a computing device that defines a rule for natural language processing of text, annotated text is selected from a first document of a plurality of annotated documents. An entity rule type is selected from a plurality of entity rule types. An argument of the selected entity rule type is identified. A value for the identified argument is randomly selected based on the selected annotated text to generate a rule instance. The generated rule instance is applied to remaining documents of the plurality of annotated documents. A rule performance measure is computed based on application of the generated rule instance. The generated rule instance and the computed rule performance measure are stored for application to other documents.

Rule development for natural language processing of text

Systems, methods and articles of manufacture are disclosed for synchronizing a transaction profile with a resolution status of a problem experienced by an application. The problem may be detected for the application. A transaction profile may be retrieved for the detected problem. The transaction profile may include a sequence of transactions to be performed on the system to remedy the open problem. Transactions occurring on the system may be monitored, and an instance of the transaction profile may be updated accordingly to create a synchronized transaction profile.

Synchronizing problem resolution task status using awareness of current state and transaction history

An approach is provided to display a common display page on a variety of display devices using the item's tab order property to determine the display order. Controls that are displayed on a display device typically have a tab order. When a window that has controls is displayed, the cursor is typically placed at the control with the lowest (i.e., first) tab order. When the user presses the tab key, the cursor moves to the control corresponding to the next lowest tab order. This allows the designer to design a single window (or panel) that is displayed differently on constrained devices. However, even though the window is displayed differently, using tab order mapping maintains a consistent visible proximity between controls despite the type of display device being used by the user.

Tab order mapping of user interfaces

A system, method, and article of manufacture are disclosed for monitoring and resolving problems detected in the application stack. The application stack may include multiple, interpedently application components which collectively provide a unified service. An interactive problem resolution program may monitor and assist users in troubleshooting an application stack installed on a separate computer system. Generally, when a problem in the application stack is detected, the IPR Program may alert users to the problem and provide information about the problem to guide users in taking steps to correct the problem.

Interactive problem resolution presented within the context of major observable application behaviors

Growing community of inventors

Cary, NC, United States of America

David B Styles