International Business Machines Corporation

Zhan Peng Huo

Lu Yan Li

Jiu Chang Du

Peng Hui Jiang

Jia Yu

Fei Tan

Xing Xing Shen

Jia Tian Zhong

Jian Ding

Xiao Ling Chen

Jia Liu

Fei Fei Li

Dong Hui Liu

Li Cao

Jiang Yi Liu

Wen Ji Huang

WeiFeng Zhang

Zhi Li Guan

Yong Yin

Xuan Feng

Han Su

Xiao Xuan Fu

Ren Jie Feng

Zhong Hao Wang

Qi Cheng Li

Cheng Fang Wang

Hui Zhao

Yan Huang

Wei Cui

Yang Liang

Si Er Han

Xue Yong Zhang

Ting Yin

Qi Li

Ying Yan

Min Cheng

Hong Wei Sun

Yu Sun

Kuo-Liang Chou

Wen Bin Han

Yong Xu

Qing Yu Pei

Kui Zhang

Shu Jun Tang

Jun Hao

Hong Yi Zhang

Pei Ci Fang

Nan Mu

Da Long Wang

Yan Fei Qin

Yu Zui You

Dian Bj Wang

Lin Lin Ll Su

Processing within a container-based computing environment is facilitated by standardizing container identification processing. The process includes, for a container being provided from a container image, extracting multiple layer files from the container image and using natural language processing to determine frequency of occurrence of software components in the multiple layer files of the container image. Further, the process includes generating a container name for the container using the determined frequency of occurrence of software components in the multiple layer files of the container image. The method further includes providing the container for the container-based computing environment using the container image and the generated container name. The generated container name facilitates processing within the container-based computing environment by being based, at least in part, on the frequency of occurrence of software components in the multiple layer files of the container image.

Container name identification processing

Computer-implemented methods for performing automated pre-processing of data for a machine-learning based prediction system are provided. Aspects include receiving a plurality of raw data sets, receiving a plurality of processed data sets, wherein each of the plurality of processed data sets corresponds to one of the plurality of raw data sets, and generating a plurality of pre-processing templates based on the plurality of raw data sets and the processed data set. Aspects also include receiving an input data set, generating, for each of the plurality of pre-processing templates, a matching score for the input data set, and selecting one of the plurality of pre-processing templates based on the matching score. Aspects further include pre-processing the input data set using the selected pre-processing template and providing the pre-processed input data set to the machine learning based prediction system.

Automated data pre-processing for machine learning

In a method for improving converter and compiler operator granularity, a processor extracts an operator granularity from an artificial intelligence framework and an original model. A processor receives device characteristics from a target device. A processor outputs a converter granularity level to a converter based on the operator granularity and the device characteristics. A processor outputs a compiler granularity level to a compiler based on the operator granularity and the device characteristics.

Optimizing operator granularity in compiling and converting artificial intelligence (AI) models

A method, system, and computer program product for dynamic cue generation for language learning are provided. The method identifies one or more listener actions for a listener engaged in listening learning. Based on the one or more listener actions, a language level of the listener is determined. A set of prompt metrics are generated based on the language level and the one or more listener actions. A set of audio stream levels is generated based on the set of prompt metrics. The method selects an audio stream level for a listening prompt, where the audio stream level is selected from the set of audio stream levels based on the language level of the listener.

Dynamic cue generation for language learning

A computer-implemented method for managing information during a web conference is provided. The computer-implemented method includes collecting and formatting meeting application information at a processor of a computing device having a screen being shared with attendees of the web conference and collecting and formatting to-be-popup application information at the processor. The computer-implemented method further includes analyzing, by the processor, an urgency of the to-be-popup application information and a correlation between the to-be-popup application information and the meeting application information and determining, by the processor, whether to share the to-be-popup application information with a user of the computing device and with the attendees based on results of the analyzing.

Dynamically controlling foreground applications during screen sharing conferences

Aspects of the invention include converting an artificial intelligence (AI) model generated in a first framework to a uniform exchange formatted model by engaging a master table to retrieve instructions for converting from the AI model to the uniform exchange formatted model in accordance with the first framework. The uniform exchange formatted model in compiled by engaging the master table to retrieve instructions for compiling the uniform exchange formatted model in accordance with the first framework. Data is received as an input to the compiled uniform exchange formatted model and an output is generated by engaging the master table to retrieve instructions for generating the output in accordance with the first framework.

Uniform artificial intelligence model conversion

A computer implemented method manages an image. A number of processor units identifies a set of base files in a set of image layers for a candidate file in a container layer in response to a request to create a new image with the candidate file. The number of processor units identifies delta data between the candidate file and the set of base files. The number of processor units creates the new image with a new top image layer using the delta data identified for the candidate file in the container layer. According to other illustrative embodiments, a computer system, and a computer program product for managing an image are provided. As a result, the illustrative embodiments provide a technical effect of reducing the size of images and reducing storage space used to store the images.

Controlling layers in container images to reduce redundant content between layers

Various embodiments of the present invention provide a method, system, and program product that perform the following operations: organizing a project having source code into a stacked-layer framework, the stacked-layer framework having: (i) a code layer including the source code of the project, (ii) a label layer including one or more label pages labelling the source code according to respective features of the source code, and (iii) a mask layer including one or more mask pages concealing portions of the source code according to labels of the label layer; receiving a request from a user to utilize a portion of the project, wherein the request includes a user requirement identifying one or more desired features; identifying a mask page of the one or more mask pages corresponding to the one or more desired features identified by the user requirement; and communicating the identified mask page to the user.

Delivering portions of source code based on a stacked-layer framework

A method, computer program product and system are provided for preloading debug information based on the presence of incremental source code files. Based on parsed input parameters to a source code debugger, a source code repository and a local storage area are searched for an incremental file. In response to the incremental file being located, a preload indicator in the incremental file, which is a source code file, is set. Based on the preload indicator being set, debug symbol data from the incremental file is merged to a preload symbol list. In response to receiving a command to examine the debug symbol data from the incremental file, the preload symbol list is searched for the requested debug symbol data.

Preloading debug information based on the increment of application

Methods, systems, and computer program products for container image management are disclosed. In a method, a first group of operations that are performed in respective layers in a base image are obtained. A second group of operations that are performed in respective layers not comprised in the base image are obtained. The second group of operations are optimized based on a comparison between the first and second groups of operations. A destination container image is generated based on the optimized second group of operations and the base image.

Container image management

A connection request from one computing entity is obtained. The connection request is a request to connect to another computing entity and includes a requested connectivity protocol configuration. A determination is made as to whether the requested connectivity protocol configuration is a match to a particular connectivity protocol configuration of the other computing entity. Based on determining that the requested connectivity protocol configuration is not a match to the particular connectivity protocol configuration of the other computing entity, a selected configuration file that is a match to the requested connectivity protocol configuration is selected from a plurality of configuration files. Based on selecting the selected configuration file, routing of the connection request to a connectivity protocol service corresponding to the selected configuration file is initiated to facilitate establishing a connection between the one computing entity and the other computing entity.

Multi-version compatibility for connectivity protocols

Intelligent layer control for building image files into new container images while avoiding redundant files stored between read-write and read-only image layers. Pre-commit processing module (PPM) filters updated files. Hash values of updated files are compared with hash values of original files. Where hash values are the same, PPM generates an OMD file that updates the inode metadata of original files with metadata of the updated file and removes the updated file. Hash values that are not the same, with differ values less than a threshold value, PPM generates an OMD file with updates to the inode metadata, content change flag and/or delta content info reflecting the updated file and removes updated file from container file system. OMD files are passed to storage driver whereby metadata and/or changes to content are overridden and merged with original file content generating a single unified user view of the container file system.

Intelligent layer control of redundant content in container images

Embodiments are for using sequential set index to determine the root cause location of software issues and problem verification of the software issues. Fail data for a software program is received. The fail data corresponds to a plurality of failed test cases for the software program. It is determined that at least one model in a plurality of models includes at least one case that matches the fail data of the software, each of the plurality of models having one or more cases, the at least one case being in the one or more cases. It is determined that a root cause of the at least one model corresponds to the plurality of failed test cases for the software program.

Use sequential set index for root cause location and problem verification

A method and system for determining causes. The method includes generating a set of cause indices. Each cause index includes a cause, at least one pattern, and a probability of the cause for the pattern. The method also includes receiving a data set, and determining that one or more indexed patterns match a pattern in the data set. Further, the method includes retrieving causes for the indexed patterns that match the pattern in the data set, as well as determining probabilities of the retrieved causes for the pattern in the data set.

Determining causes of events in data

Growing community of inventors

Beijing, China

Zhan Peng Huo