International Business Machines Corporation

Nuance Communications, Inc.

Bin Jia

Richard R Treumann

Ying Liu

Yan Rong Shen

Zhen Zhang

Jia Zhong Wu

E Feng Lu

Robert Sidney Blackmore

Shuai Wang

Hong Wei Sun

Guo Kang Fu

James Robert Magowan

Guy Patrick Hindle

Xi Xue Jia

Zhao Han

Ellick C Law

Callum Peter Jackson

Philip L Norton

Matthew W Leming

Eric William Janke

Laura Jane Cowen

Mark R Wilson

Anne Marie Workman

An example operation may include one or more of receiving a plurality of transactions by an agent node, selecting, by the agent node, transactions from the plurality of the transactions that are the most likely to be conflicting based on common factors, combining, by the agent node, the selected transactions into a batch, and reordering the selected transactions within the batch to increase a success rate.

Batch processing

An example operation may include one or more of submitting, by a blockchain client, a transaction proposal to a blockchain network, endorsing, by one or more endorser nodes or peers, the transaction proposal, endorsing comprising updating an in-process world state list, providing one or more composite endorsements, each comprising a base result and a forecasted result corresponding to the transaction proposal, submitting an endorsed transaction in response to the one or more composite endorsements, validating, by a committer node or peer, the endorsed transaction, committing the endorsed transaction to the blockchain network, and removing the committed transaction from the in-process world state list.

Database composite endorsement

Flexible utilization of capacity to provide computing services mediated by a service broker. The service broker receives a request from a first party for assistance with a service workload as well as an indication from a second party of a willingness to provide service. The service broker matches the parties with each other based at least in part on the capacity of the second party to service the service workload of the first party, manages offloading of the service workload from first party to the second party, and manages billing of a workload provider (such as the first party or the first party's customer) for the offloading engagement.

Service broker for computational offloading and improved resource utilization

A method, program product and a system is provided for managing a cache. The method includes analyzing at least an intermediate product of a computer program. The intermediate product is produced by the computer program in response to a set of control inputs. The method also includes determining a resource measure associated with the first intermediate product and determining a resource measure value for the first intermediate product using a first set of control inputs&#x3e; The first intermediate product is stored in the cache upon determination that the resource measure value exceeds a predetermined resource threshold.

Managing a cache for storing one or more intermediate products of a computer program

According to embodiments, method, systems, and computer program products are provided for a web server transmitting a response to an access request. An access request for a web application program deployed on a web server is received, where a response corresponding to the access request is composed with a script. A test script is transmitted. A response time of the test script is received. The response time includes a time that the test script is executed in a browser of a client and a time that the test script is transferred over a network. Different responses are transmitted to different types of scripts included in the response corresponding to the access request in accordance with the response time of the test script.

Method and system for a web server transmitting a response to an access request

Maintaining and supplying a plurality of speech models is provided. A plurality of speech models and metadata for each speech model are stored. A query for a speech model is received from a source. The query includes one or more conditions. The speech model with metadata most closely matching the supplied one or more conditions is determined. The determined speech model is provided to the source. A refined speech model is received from the source, and the refined speech model is stored.

Maintaining and supplying speech models

In an interconnection network, multiple nodes are connected to one of a first layer of switches. The first layer of switches is connected to one another through a second layer of switches. Each of the nodes is connected through one of multiple shared links connecting the first layer switches and the second layer of switches. A pipelined broadcast manager schedules broadcast steps for a pipelined broadcast through the crossbar switches according to each separate, single static route set through at least one of an inter-switch broadcast phase during which at least one packet is passed from at least one switch of the crossbar switches to another of the crossbar switches and an intra-switch broadcast phase during which the at least one packet is passed only among a selection of nodes connected to the another of the crossbar switches.

Contention free pipelined broadcasting within a constant bisection bandwidth network topology

A method for processing a data object for a database, the database containing data representing a first data model and a set of one or more mapping rules, includes receiving a data object that conforms to a second data model. The method then selects one or more of the mapping rules. The mapping rules provide a mapping between a set of elements of the second data model and a corresponding set of elements of the first data model. The method applies the selected mapping rules to transform a set of elements of the received data object into a corresponding set of elements of a target data object conforming to the first data model. The method then searches the database for the set of elements of the target data object to identify instances of the target data object in the database. A corresponding computer program product and apparatus are also disclosed.

Avoiding duplicate and equivalent data objects in service registries

The present invention discloses converting a text form into a speech. In the present invention, partial word lists of a data source are obtained by parsing the data source in parallel or in series. The partial word lists are then compiled to obtain phoneme graphs corresponding, respectively, to the partial word lists, and then the obtained phoneme graphs are combined. Speech recognition is then conducted according to the combination results. According to the present invention, computational complexity may be reduced and recognition efficiency may be improved during speech recognition.

Partial word lists into a phoneme tree

Converting partial word lists into a phoneme tree for speech recognition

In an interconnection network, multiple nodes are connected to one of a first layer of switches. The first layer of switches is connected to one another through a second layer of switches. Each of the nodes is connected through one of multiple shared links connecting the first layer switches and the second layer of switches. A pipelined broadcast manager schedules a hierarchical pipelined broadcast through at least one switch of the first layer switches comprising non-root nodes by selecting two nodes among the non-root nodes connected to the at least one switch and scheduling each of multiple broadcast steps for the pipelined broadcast with at least one of an inter-switch broadcast phase and an intra-switch broadcast phase using the selected two nodes.

Distributing a computing operation among processes and for gathering results of the computing operation from the plurality of processes. An exemplary method includes the operations of pairing a plurality of processes such that each process has a maximum of one interaction partner, selecting half of the data located at a process, dividing the selected half of the data into a plurality of data segments, transmitting a first data segment resulting from the dividing operation from the process to the interaction partner of the process, receiving a second data segment at the process from the interaction partner, concurrently with the transferring and receiving operations, performing a computing operation on a third data segment previously received from a previous interaction partner and a fourth data segment from the data segments, and iterating over the transmitting, receiving and computing operations until all the data segments have been exchanged.

Iterative process partner pairing scheme for global reduce operation

A parallel processing data processing system builds at least one data structure indicating a communication schedule for a plurality of processes each having a respective one of a plurality of equal length vectors formed of multiple equal size chunks. The data processing system, based upon the at least one data structure, communicates chunks of the plurality of vectors among the plurality of processes and performs partial reduction operations on chunks in accordance with the communication schedule. The data processing system then stores a result vector representing reduction of the plurality of vectors.

Growing community of inventors

Poughkeepsie, NY, United States of America

Bin Jia