Abstract: Speech processing tasks may be allocated at least partly to a local device (e.g., user computing device that receives spoken words) and at least partly to a remote device to determine one or more user commands or tasks to be performed by the local device. The remote device may be used to process speech that the local device could not process or understand, or for other reasons, such as for error checking. The local device may then execute or begin to execute locally determined tasks to reduce user-perceived latency. Meanwhile, the entire media input, or a portion thereof, may be sent to the remote device to process speech, verify the tasks and/or identify other user commands in the media input (or portion thereof).

Abstract: A computer implemented of evaluating market participants' stock portfolios and distributing actionable environmental, social and governance (ESG) stock data over a network to multiple remote market participant terminals allowing market participants to easily rebalance their portfolios so that their investments align with their personal values or social tolerances. The method comprises receiving from multiple sources non-standardized electronic ESG data for a plurality of securities, mapping, scaling and normalizing the data to present a final impact score that the market participant can act upon.

Abstract: Various embodiments of the present disclosure provide a multi-stage system and process for removing debris from plastic sheets (such as tier sheets or any other suitable sheets formed from any suitable material). Certain embodiments of the sheet-cleaning system (10) include a drive assembly comprising a drive roller (211, 212) driven by a drive actuator, an electrostatic-discharge device (220) operable to discharge static electricity, and a cleaning roller assembly (230) comprising a cleaning roller (231, 232) driven by a cleaning roller actuator, the cleaning roller (231, 232) comprising a cleaning implement on at least part of its outer surface.

Abstract: A computer implemented of evaluating market participants' stock portfolios and distributing actionable environmental, social and governance (ESG) stock data over a network to multiple remote market participant terminals allowing market participants to easily rebalance their portfolios so that their investments align with their personal values or social tolerances. The method comprises receiving from multiple sources non-standardized electronic ESG data for a plurality of securities, mapping, scaling and normalizing the data to present a final impact score that the market participant can act upon.

Abstract: A device receives, based on a user interaction with a chatbot, request data for a request for a service associated with a software tool that is part of a service management system. The device identifies, based on the request data, the service that is being requested and a first set of entity data for a first set of entities relating to the service. The device obtains a second set of entity data for a second set of entities by using an application programming interface (API) to reference a data structure that is associated with the service management system and that stores the second set of entity data. The device causes one or more components of the service management system to use the first set of entity data to execute a set of actions to perform the service.

Abstract: A device receives, based on a user interaction with a chatbot, request data for a request for a service associated with a software tool that is part of a service management system. The device identifies, based on the request data, the service that is being requested and a first set of entity data for a first set of entities relating to the service. The device obtains a second set of entity data for a second set of entities by using an application programming interface (API) to reference a data structure that is associated with the service management system and that stores the second set of entity data. The device causes one or more components of the service management system to use the first set of entity data to execute a set of actions to perform the service.

Abstract: Speech processing tasks may be allocated at least partly to a local device (e.g., user computing device that receives spoken words) and at least partly to a remote device to determine one or more user commands or tasks to be performed by the local device. The remote device may be used to process speech that the local device could not process or understand, or for other reasons, such as for error checking. The local device may then execute or begin to execute locally determined tasks to reduce user-perceived latency. Meanwhile, the entire media input, or a portion thereof, may be sent to the remote device to process speech, verify the tasks and/or identify other user commands in the media input (or portion thereof).

Abstract: Avatar animation may be enhanced to reflect emotion and other human traits when animated to read messages received from other users or other messages. A message may be analyzed to determine visual features associated with data in the message. The visual features may be depicted graphically by the avatar to create enhanced avatar animation. A text-based message may include indicators, such as punctuation, font, words, graphics, and/or other information, which may be extracted to create the visual features. This information may be used to select visual features as special animation, which may be implemented in animation of the avatar. Examples of visual features include animations of laugher, smiling, clapping, whistling, and/or other animations.

Abstract: Various embodiments are directed to systems and methods for executing web applications. A computing device may execute a common view process that comprises a web view. The computing device may receive a request to launch a web application. The computing device may execute a web application interface process programmed to determine a location indicator indicating a location of code for executing the web application and send the location indicator to the web view. The web view may receive the code for executing the web application. The code may be executed by the web view.

Abstract: A method for optimizing bytecode in the presence of try-catch blocks comprises generating an Intermediate Representation of the bytecode, scanning each basic block of the bytecode to identify try blocks, scanning each basic block of the bytecode to identify try-blocks, splitting each try-block into a first half and a second half at the first statement that can throw an exception, establishing an edge between the first half and the second half of each try-block, between the first half and the catch block, and between the catch block and the basic block subsequent to the second half of each try-block.

Abstract: According to a first aspect of the present invention, a method for linking bytecodes of an uninterrupted block of bytecodes in the formation of a data flow graph comprises the steps of scanning the uninterrupted block of bytecodes in a forward manner to identify the start of each of the bytecodes, scanning in a backward manner bytecodewise each of the bytecodes in the uninterrupted block of bytecodes, and generating a link in the data flow graph that links each of the bytecodes to all other of the bytecodes used by the each of the bytecodes.