Abstract: Executing a map reduce sequence may comprise executing all jobs in the sequence by a collection of a plurality of processes with each process running zero or more mappers, combiners, partitioners and reducers for each job, and transparently sharing heap state between the jobs to improve metrics associated with the job. Processes may communicate among themselves to coordinate completion of map, shuffle and reduce phases, and completion of said all jobs in the sequence.
Type:
Application
Filed:
August 24, 2012
Publication date:
February 27, 2014
Applicant:
INTERNATIONAL BUSINESS MACHINES CORPORATION
Inventors:
David Cunningham, Benjamin W. Herta, Vijay A. Saraswat, Avraham E. Shinnar
Abstract: Interactive iterative program parallelization based on dynamic feedback program parallelization, in one aspect, may identify a ranked list of one or more candidate pieces of code each with one or more source refactorings that can be applied to parallelize the code, apply at least one of the one or more refactorings to create a revised code, and determine performance data associated with the revised code. The performance data may be used to make decisions on identifying next possible ranked list of refactorings.
Type:
Application
Filed:
February 22, 2010
Publication date:
August 25, 2011
Applicant:
INTERNATIONAL BUSINESS MACHINES CORPORATION
Inventors:
Evelyn Duesterwald, Robert M. Fuhrer, Vijay Saraswat
Abstract: Executing a map reduce sequence may comprise executing all jobs in the sequence by a collection of a plurality of processes with each process running one or more mappers, combiners, partitioners and reducers for each job, and transparently sharing heap state between the jobs to improve metrics associated with the job. Processes may communicate among themselves to coordinate completion of map, shuffle and reduce phases, and completion of said all jobs in the sequence.
Type:
Application
Filed:
August 25, 2012
Publication date:
February 27, 2014
Applicant:
INTERNATIONAL BUSINESS MACHINES CORPORATION
Inventors:
David Cunningham, Benjamin W. Herta, Vijay A. Saraswat, Avraham E. Shinnar
Abstract: Executing a map reduce sequence may comprise executing all jobs in the sequence by a collection of a plurality of processes with each process running one or more mappers, combiners, partitioners and reducers for each job, and transparently sharing heap state between the jobs to improve metrics associated with the job. Processes may communicate among themselves to coordinate completion of map, shuffle and reduce phases, and completion of said all jobs in the sequence.
Type:
Grant
Filed:
August 25, 2012
Date of Patent:
September 29, 2015
Assignee:
International Business Machines Corporation
Inventors:
David Cunningham, Benjamin W. Herta, Vijay A. Saraswat, Avraham E. Shinnar
Abstract: Interactive iterative program parallelization based on dynamic feedback program parallelization, in one aspect, may identify a ranked list of one or more candidate pieces of code each with one or more source refactorings that can be applied to parallelize the code, apply at least one of the one or more refactorings to create a revised code, and determine performance data associated with the revised code. The performance data may be used to make decisions on identifying next possible ranked list of refactorings.
Type:
Grant
Filed:
February 22, 2010
Date of Patent:
May 13, 2014
Assignee:
International Business Machines Corporation
Inventors:
Evelyn Duesterwald, Robert M. Fuhrer, Vijay Saraswat
Abstract: A method, apparatus and computer program product for responding to an indirect utterance in a dialogue between a user and a conversational system is described. An indirect utterance is received. A parse structure of the indirect utterance is generated. The indirect utterance is an utterance which does not match a user goal expressed as elements of a knowledge graph. The parse structure is connected through the knowledge graph to a user goal to issue a request which is not stated in the indirect utterance. A system response is performed, where the system response is a dialogue system response based on a combination of the parse structure and the connected user goal.
Type:
Application
Filed:
February 1, 2018
Publication date:
August 1, 2019
Inventors:
Mustafa Canim, Robert G. Farrell, Achille B. Fokoue-Nkoutche, John A. Gunnels, Ryan A. Musa, Vijay A. Saraswat
Abstract: Qualitative reasoning for physical systems, and techniques for automatic code generation are used to automatically construct digital controllers/device drivers for electro-mechanical systems. Such construction uses models of the system's components described as finite state machines, to form a configuration space. Transitions of the configuration space are labelled as external or internal and a state of the system is identified as a desired state while other states are identified as undesirable. From this configuration space a controller is generated to drive the system to the desired state while avoiding the undesirable states.
Type:
Grant
Filed:
June 7, 1995
Date of Patent:
November 3, 1998
Assignee:
XEROX Corporation
Inventors:
Daniel G. Bobrow, Vijay A. Saraswat, Vineet Gupta
Abstract: A method, apparatus and computer program product for responding to an indirect utterance in a dialog between a user and a conversational system is described. An indirect utterance is received. A parse structure of the indirect utterance is generated. The indirect utterance is an utterance which does not match a user goal expressed as elements of a knowledge graph. The parse structure is connected through the knowledge graph to a user goal to issue a request which is not stated in the indirect utterance. A system response is performed, where the system response is a dialog system response based on a combination of the parse structure and the connected user goal.
Type:
Grant
Filed:
February 1, 2018
Date of Patent:
January 14, 2020
Assignee:
International Business Machines Corporation
Inventors:
Mustafa Canim, Robert G Farrell, Achille B Fokoue-Nkoutche, John A Gunnels, Ryan A Musa, Vijay A Saraswat
Abstract: The example computer-implemented method may comprise computing, by a generator processor on each of a plurality of learners, a gradient for a mini-batch using a current weight at each of the plurality of learners. The method may also comprise generating, by the generator processor on each of the plurality of learners, a plurality of triples, wherein each of the triples comprises the gradient, the weight index of the current weights used to compute the gradient, and a mass of the gradient. The method may further comprise performing, by a reconciler processor on each of the plurality of learners, an allreduce operation on the plurality of triples to obtain an allreduced triple sequence. Additionally, the method may comprise updating, by the reconciler processor on each of the plurality of learners, the current weight at each of the plurality of learners to a new current weight using the allreduced triple sequence.
Abstract: A method and apparatus provide a service for permitting parties to engage in instant messaging and to allow parties so engaged to initiate a separate telephone call which may be billed. According to one method a server, referred to as a Phone Web server acts as an interface between a computer user and a telephone user. The interface is coupled to an instant messaging platform enabling the two users to exchange instant messages. One user can prompt the other to suggest the initiation of a telephone call. If the other party accepts the suggestion then the Phone Web server brokers a phone call between the two parties. In a variation of this embodiment both users could be communicating via computers and can launch the telephone initiation process.
Abstract: A method for increasing performance of an operation on a distributed memory machine is provided. Asynchronous parallel steps in the operation are transformed into synchronous parallel steps. The synchronous parallel steps of the operation are rearranged to generate an altered operation that schedules memory accesses for increasing locality of reference. The altered operation that schedules memory accesses for increasing locality of reference is mapped onto the distributed memory machine. Then, the altered operation is executed on the distributed memory machine to simulate local memory accesses with virtual threads to check cache performance within each node of the distributed memory machine.
Type:
Application
Filed:
November 12, 2010
Publication date:
May 17, 2012
Applicant:
INTERNATIONAL BUSINESS MACHINES CORPORATION
Inventors:
George Almasi, Guojing Cong, David J. Klepacki, Vijay A. Saraswat
Abstract: The present information processing apparatus executes fault-tolerant distributed computing through a network and the information processing apparatus comprises: a finish initiator that creates a finish state on a home node; an activity creator that instructs to create at least one activity on a remote node; an activity launcher that allows to store in a resilient store a value indicating that at least one live activity is present on the remote node; an activity closer that decreases the number in the local memory when the living activity is completed and resets the value in the resilient store to indicate that no live activity is present on the remote node; and a finish closer that closes the finish state when the value stored in the resilient store indicates an absence of remote nodes having at least one live activity.
Abstract: Automated feature construction for algorithm portfolios in machine learning is provided. A gray scale image is generated from a text representing a problem instance. The gray scale image is rescaled or reshaped to a predefined size that is smaller than an initial size of the gray scale image. The rescaled gray scale image represents features of the problem instance. The rescaled gray scale image is input as features to a machine learning-based convolutional neural network. Based on the rescaled gray scale image, the machine learning-based convolutional neural network is automatically trained to learn to automatically determine one or more problem solvers from a portfolio of problem solvers suited for solving the problem instance.
Type:
Grant
Filed:
February 4, 2016
Date of Patent:
January 17, 2017
Assignee:
International Business Machines Corporation
Inventors:
Andrea Loreggia, Yuri Malitsky, Horst C. Samulowitz, Vijay A. Saraswat
Abstract: A method for increasing performance of an operation on a distributed memory machine is provided. Asynchronous parallel steps in the operation are transformed into synchronous parallel steps. The synchronous parallel steps of the operation are rearranged to generate an altered operation that schedules memory accesses for increasing locality of reference. The altered operation that schedules memory accesses for increasing locality of reference is mapped onto the distributed memory machine. Then, the altered operation is executed on the distributed memory machine to simulate local memory accesses with virtual threads to check cache performance within each node of the distributed memory machine.
Type:
Grant
Filed:
November 12, 2010
Date of Patent:
October 21, 2014
Assignee:
International Business Machines Corporation
Inventors:
George Almasi, Guojing Cong, David J. Klepacki, Vijay A. Saraswat
Abstract: The example computer-implemented method may comprise computing, by a generator processor on each of a plurality of learners, a gradient for a mini-batch using a current weight at each of the plurality of learners. The method may also comprise generating, by the generator processor on each of the plurality of learners, a plurality of triples, wherein each of the triples comprises the gradient, the weight index of the current weights used to compute the gradient, and a mass of the gradient. The method may further comprise performing, by a reconciler processor on each of the plurality of learners, an allreduce operation on the plurality of triples to obtain an allreduced triple sequence. Additionally, the method may comprise updating, by the reconciler processor on each of the plurality of learners, the current weight at each of the plurality of learners to a new current weight using the allreduced triple sequence.
Type:
Grant
Filed:
May 5, 2016
Date of Patent:
April 21, 2020
Assignee:
INTERNATIONAL BUSINESS MACHINES CORPORATION
Abstract: Running a global production rule on data distributed over a plurality of machines may comprise receiving a local production rule that can run on each of the plurality of machines to jointly accomplish a global computation specified by the global production rule. The local production rule may be deployed to each of the plurality of machines, each of which stores a portion of the data and runs an instance of a rules engine that can run the local production rule. The plurality of machines are enabled to communicate intermediate data produced by the instance of the rules engine running the local production rule on said each of the machines. Coordinating between the plurality of machines is enabled to synchronize one or more local computations performed locally according to the local production rule on said each machine.
Type:
Application
Filed:
December 13, 2016
Publication date:
March 30, 2017
Inventors:
Matthew R. Arnold, Martin J. Hirzel, Vijay A. Saraswat, Avraham E. Shinnar, Jerome Simeon, Lionel A. Villard
Abstract: The present information processing apparatus executes fault-tolerant distributed computing through a network and the information processing apparatus comprises: a finish initiator that creates a finish state on a home node; an activity creator that instructs to create at least one activity on a remote node; an activity launcher that allows to store in a resilient store a value indicating that at least one live activity is present on the remote node; an activity closer that decreases the number in the local memory when the living activity is completed and resets the value in the resilient store to indicate that no live activity is present on the remote node; and a finish closer that closes the finish state when the value stored in the resilient store indicates an absence of remote nodes having at least one live activity.
Type:
Grant
Filed:
February 20, 2015
Date of Patent:
January 2, 2018
Assignee:
International Business Machines Corporation
Abstract: A method, apparatus and computer program product for responding to an indirect utterance in a dialogue between a user and a conversational system is described. An indirect utterance is received. A parse structure of the indirect utterance is generated. The indirect utterance is an utterance which does not match a user goal expressed as elements of a knowledge graph. The parse structure is connected through the knowledge graph to a user goal to issue a user request which is not stated in the indirect utterance. The parse structure is connected using a matching process which matches the parse structure with the connected user goal in the knowledge graph according to a similarity of the parse structure and a portion of the knowledge graph including the connected user goal. A system response is performed based on the connected user goal.
Type:
Application
Filed:
November 7, 2019
Publication date:
March 5, 2020
Inventors:
Mustafa Canim, Robert G. Farrell, Achille B. Fokoue-Nkoutche, John A. Gunnels, Ryan A. Musa, Vijay A. Saraswat
Abstract: A system dynamically schedules the creation of images in an on-demand printing apparatus outputting simplex and duplex prints. For each simplex print, a simplex print block is appended to a schedule of images to be created, and for each duplex print, a duplex print block is appended to the schedule. The duplex block includes two portions, one portion for each image in the duplex print, separated by a number of pitches which can be occupied by other images. The duplex and simplex blocks are then organized in a manner which preserves the desired final sequence of prints and minimizes the number of unused image pitches in the printing apparatus. The scheduling technique can also be applied to multi-pass color printing.
Type:
Grant
Filed:
April 21, 1995
Date of Patent:
April 2, 1996
Assignee:
Xerox Corporation
Inventors:
Vijay A. Saraswat, Daniel G. Bobrow, Markus P. J. Fromherz, Timothy G. Lindholm, Pierre C. Berlandier, John H. Conley
Abstract: A method, apparatus and computer program product for responding to an indirect utterance in a dialogue between a user and a conversational system is described. An indirect utterance is received. A parse structure of the indirect utterance is generated. The indirect utterance is an utterance which does not match a user goal expressed as elements of a knowledge graph. The parse structure is connected through the knowledge graph to a user goal to issue a user request which is not stated in the indirect utterance. The parse structure is connected using a matching process which matches the parse structure with the connected user goal in the knowledge graph according to a similarity of the parse structure and a portion of the knowledge graph including the connected user goal. A system response is performed based on the connected user goal.
Type:
Grant
Filed:
November 7, 2019
Date of Patent:
November 10, 2020
Assignee:
International Business Machines Corporation
Inventors:
Mustafa Canim, Robert G Farrell, Achille B Fokoue-Nkoutche, John A Gunnels, Ryan A Musa, Vijay A Saraswat