International Business Machines Corporation

Jian F Zhang

Yue Wang

Da Li Liu

Jing Xing

Qi Feng Huo

Dong Jun Zong

Peng Hui Jiang

Xing Fu

Hai F Zhang

Jing Lai

Su Liu

Yong Wang

Michael Baessler

Juan Liu

Hang Xiao

Yan Jia

Rongsong Shen

Yanmin Jia

Cognitive determination of whether a message is suitable for sending over a data communications network can include extracting tokens from the message prior to transmitting the message. One or more intended recipients of the message can be determined from the tokens. A machine learning classification model corresponding to the one or more recipients of the message can be selected. The machine learning classification model can be constructed based on tokens extracted from prior messages, which are combined to create a plurality of documents for training the machine learning classification model. The one or more tokens extracted from the message can be classified using the machine learning classification model. An alert message can be generated in response to determining based on the classifying that the message is unsuited for sending.

Cognitive determination of message suitability

Edge device task management by receiving an indicator corresponding to a first container running a task on a first edge device of a cluster of edge devices, wherein the indicator indicates an error status of the first container, and wherein task data of the task is stored in a first local storage of the first edge device, selecting a second edge device from the cluster of edge devices, wherein a second container on the second edge device is to run the task, instructing the first and second edge devices to transfer the task data from the first local storage of the first edge device to a second local storage of the second edge device, and in response to receiving a notification that indicates the task data has been transferred from the first local storage to the second local storage, sending the task to the second container.

Transferring task data between edge devices in edge computing

Embodiments generally enable a mobile device to display a window of a remote desktop on a mobile device with a native layout. In some embodiments, a method includes receiving a remote desktop display request from a mobile client device, wherein the remote desktop display request includes display information of the mobile client device. The method further includes generating a copy of a window process of a remote desktop computer. The method further includes generating a virtual display based at least in part on the copy of the window process of the remote desktop computer and on the display information of the mobile client device. The method further includes sending virtual display information to the mobile client device based at least in part on the virtual display.

Displaying a window of a remote desktop computer on a mobile device with a native layout

An injection system and method for monitoring and controlling task execution in a distributed computer system. The distributed computing environment is comprised of a plurality of computing nodes for executing an application, wherein the application is divided into tasks that are executed within containers on each of the nodes, and the tasks process vectors comprised of one or more data elements. A control plug-in injected into a specified container for execution with the tasks, wherein the control plug-in profiles footprint information for a targeted vector operated on by the tasks executed within the specified container.

Injection method of monitoring and controlling task execution in a distributed computer system

Provided are techniques for barcode processing. A barcode is detected. Then the barcode is classified as any of a first class code that needs to be preserved and a second class code that does not need to be preserved. For the barcode classified as the first class code, a content referred to by the barcode is saved.

Barcode processing

Embodiments of the present invention relate to a font match technology, and in particular relate to a technology for performing font match with attribute values of a font. A method for font matching may include generating a two-dimensional link matrix based on available fonts and attribute values of the available fonts, receiving a font match query, and determining a font matching the received font match query from the available fonts by using the two-dimensional link matrix. Font matching may be implemented automatically, quickly, and accurately, without the need of manually randomly trying each font of a plurality of available fonts. Meanwhile, this method is independent from text processing software and may be called by any arbitrary text processing software (application). Moreover, this method allows the user to flexibly configure selection of the font attributes so as to perform font match more efficiently.

Quick font match

Growing community of inventors

Beijing, China

Jian F Zhang