International Business Machines Corporation

Harvard College

David A Cox

Fabian F Morgan

Samuel English Anthony

Danny Yen-Fu Chen

Sheryl S Kinstler

Walter Scheirer

Ken Nakayama

Danny Y Chen

Brent Russel Phillips

John H Smith

Chuang Gan

Yang Zhang

Shiyu Chang

Kaizhi Qian

A method, a structure, and a computer system for decomposing speech. The exemplary embodiments may include one or more encoders for generating one or more encodings of a speech input comprising rhythm information, pitch information, timbre information, and content information, and a decoder for decoding the one or more encodings.

Unsupervised speech decomposition

In various embodiments, training objects are classified by human annotators, psychometric data characterizing the annotation of the training objects is acquired, a human-weighted loss function based at least in part on the classification data and the psychometric data is computationally derived, and one or more features of a query object are computationally classifies based at least in part on the human-weighted loss function.

Systems and methods for machine learning enhanced by human measurements

Systems and methods for allocating computing resources for a multi-tiered application are disclosed. A computer-implemented method includes: determining, by a computing device, a topology of a multi-tiered application; determining, by the computing device, a modeled setting of a computing resource for the multi-tiered application based on the determined topology; determining, by the computing device, an actual usage of the computing resource by the multi-tiered application; and adjusting, by the computing device, an allocation of the computing resource to the multi-tiered application based on the actual usage and the modeled setting.

Assessment and dynamic provisioning of computing resources for multi-tiered application

In various embodiments, training objects are classified by human annotators, psychometric data characterizing the annotation of the training objects is acquired, a human-weighted loss function based at least in part on the classification data and the psychometric data is computationally derived, and one or more features of a query object are computationally classified based at least its part on the human-weighted loss function.

In various embodiments, training objects are classified by human annotators, psychometric data characterizing the annotation of the training objects is acquired, a human-weighted loss function based at least in part on the classification data and the psychometric data is computationally derived, and one or more features of a query object are computationally classified based at least in part on the human-weighted loss function.

A system, a method, and a computer readable media for identifying a user-initiated log file record in a log file are provided. The log file has a user-initiated log file record and a repeating pattern of log file records automatically generated by a software program. The system allows a user to identify first and second timestamp values corresponding to first and second times which identify a time interval of interest in the log file. The system further analyzes the log file to identify the user-initiated log file record having a timestamp value between the first and second timestamp values. The system further identifies the repeating pattern of log file records in the log file.

System method, and computer readable media for identifying a user-initiated log file record in a log file

System, method, and computer readable media for identifying a user-initiated log file record in a log file

A system, a method, and a computer readable media for identifying a log file record in a log file are provided. The log file has a plurality of log file records therein. The plurality of log file records has a repeating pattern of log file records. The method includes displaying a first graphical user interface on a display device. The method further includes selecting the log file utilizing the first graphical user interface. The method further includes analyzing the log file to determine the repeating pattern of log file records. The method further includes selecting a first command button on the first graphical user interface that executes an algorithm to analyze additional log file records that were automatically added to the log file by a software program. The additional log file records are associated with the repeating pattern of log file records. The method further includes storing the first log file record in a memory device if a first log file record in the additional log file records has at least one command-related attribute that does not correspond to command-related attributes in the repeating pattern of log file records.

System, method, and computer readable media for identifying a log file record in a log file

An XML Data Loader which employs the SAX parsing technology, provides an extendible architecture to generate SQL for each type of XML document to be loaded, and leverages multithreading and decoupling of processes to parse an XML file and to update a database. By using object-oriented programming methodologies, the system and method create new instances of the XML data loader for each XML file to be loaded into a database. Thus, many XML data files may be processed and loaded simultaneously, which minimizes system memory requirements, improves system reliability and memory management, and reduces processing time required from the start of processing an XML file to completing its loading into a database. The invention is applicable to other types of markup language documents, as well.

Growing community of inventors

Sugar Grove, IL, United States of America

David A Cox