International Business Machines Corporation

Harini Srinivasan

Sophia Krasikov

Jong-Deok Choi

Aditya Vempaty

Ganesan Ramalingam

Jason Boada

Kavitha Srinivas

Eoin Lane

Peter Francis Sweeney

Michael Desmond

Nicholas Matthew Mitchell

Harold Leon Ossher

Michael George Burke

Edith Gail Schonberg

Ravi B Konuru

Gary Seth Sevitsky

Matthew J Callery

James Conallen

Theodore Cheuk-Tak Law

A method for generating an output comprising one or more segments includes obtaining a plurality of profiles derived from unstructured data associated with a plurality of users, wherein a given one of the profiles corresponds to a respective one of the users; repetitively executing at least one machine learning technique on the plurality of profiles, each execution producing a respective set of one or more segments from the plurality of profiles; generating a complete graph by performing pairwise comparisons between sets of segments from respective executions; applying at least one persistency graph algorithm to the complete graph to find one or more coherent clusters comprising one or more segments that are persistent across the repetitive executions of the machine learning technique; and producing the output at least in part by selecting at least one of the segments from at least one of the coherent clusters.

Automatic segmentation of a collection of user profiles

Techniques are presented for providing a software fitting assessment. The techniques may be performed by methods, apparatus, and/or computer program products. The techniques include automatically matching on a computer system one or more specified requirements for a project with one or more software functions stored in a repository. The automatically matching includes mining the repository in order to match requirements. The repository includes software functions, requirements accumulated from previous projects, and results of stored matches between the software functions and the requirements accumulated from previous projects. The techniques include outputting by the computer system one or more results of the matching.

Techniques providing a software fitting assessment

A method (and system) of modeling runtime behavior for a software application includes limiting a portion of a run to be analyzed using criteria that define an analysis scenario, structuring events of the analysis scenario as a flow of logical content through a plurality of transformations, applying labels to elements of the flow, and generating a behavior signature for the analysis scenario based on the labels.

Method and system for modeling runtime behavior

The present invention discloses a software development tool that includes a model development interface, an aspect design tool, and a model transformation engine. The model development interface can permit a developer to graphically design at least one software model. The aspect design tool for the model development interface can be configured to permit a user to specify at least one aspect to be applied to a model of the model development interface. The model transformation engine can transform a model of the model development interface having at least one aspect defined via the aspect design tool into automatically generated aspect code integrated with base language code.

Model driven development including aspect integration tool

A process of analyzing performance of a computer program including a plurality of components includes steps of: receiving a call tree profile having a plurality of nodes, each node representing an invocation and each node being annotated with a cumulative performance cost for the node; a set of one or more components of interest to a user; and a threshold level of a cumulative performance cost associated with invocations of methods of the component that constitutes an expensive method invocation; analyzing the call tree profile from leaf nodes, the analyzing step comprising selecting a set of one or more components of interest to a user and selecting only invocations that exceed the threshold level, and pruning the set of invocations that are not selected. Finally the process includes a step of presenting to the user a list of costs associated with the selected invocations.

Summarizing application performance in a large system from a components perspective

A method for recording and replaying execution of distributed programs on a computer system in a distributed environment, includes identifying an execution order of critical events of a program, generating groups of critical events of the program, wherein for each group, critical events belonging to the group belong to a common execution thread, and generating for each execution thread a logical thread schedule that identifies a sequence of the groups so as to allow deterministically replaying a non-deterministic arrival of stream socket connection requests, a non-deterministic number of bytes received during message reads, and a non-deterministic binding of stream sockets to local ports.

Method and system for recording and replaying the execution of distributed java programs

The mechanism includes three data structures: object data associated with each object O1, O2, O3 ... of a program, a class descriptor associated with the classes to which the objects O1, O2, O3 ... belong, and interface references. For each class C.sub.i, the class descriptor corresponding to the class C.sub.i includes: i) a class method table comprising one or more entries, wherein each entry identifies an implementation of a method inherited or defined by the class C.sub.i ; and ii) for each interface implemented by the class C.sub.i, an interface method table comprising one or more entries, wherein each entry identifies an implementation of a method declared by the particular interface or an implementation of a method declared by an interface extended by the particular interface. An interface reference for an interface declaration ID associated with a given object OI, wherein the object OI implements methods that are declared by an interface declaration ID includes a pointer to the interface method table corresponding to the interface declaration ID within the class descriptor corresponding to the object OI. An efficient mechanism for method invocations is provided by generating an index that identifies the entry corresponding to the appropriate method in the interface method table pointed to by the pointer of the interface reference, reading the data stored by the entry identified by the index, and controlling the computer processing apparatus to jump to the address identified by the data.

Object model for Java.TM.

A multithreaded program includes sequences of events wherein each sequence is associated with one of a plurality of execution threads. In a record mode, the software tool of the present invention records a run-time representation of the program by distinguishing critical events from non-critical events of the program and identifying the execution order of such critical events. Groups of critical events are generated wherein, for each group G.sub.i, critical events belonging to the group G.sub.i belong to a common execution thread, critical events belonging to the group G.sub.i are consecutive, and only non-critical events occur between any two consecutive critical events in the group G.sub.i. In addition, the groups are ordered and no two adjacent groups include critical events that belong to a common execution thread. For each execution thread, a logical thread schedule is generated that identifies a sequence of said groups associated with the execution thread. The logical thread schedules are stored in persistent storage for subsequent reuse. In a replay mode, for each execution thread, the logical thread schedule associated with the execution thread is loaded from persistent storage and the critical events identified by the logical thread schedule are executed.

Deterministic replay of multithreaded applications

The mechanism includes three data structures: object data associated with each object O1, O2, O3 . . . of a program, a class descriptor associated with the classes to which the objects O1, O2, O3 . . . belong, and interface references. For each class C.sub.i, the class descriptor corresponding to the class C.sub.i includes: i) a class method table comprising one or more entries, wherein each entry identifies an implementation of a method inherited or defined by the class C.sub.i ; and ii) for each interface implemented by the class C.sub.i, an interface method table comprising one or more entries, wherein each entry identifies an implementation of a method declared by the particular interface or an implementation of a method declared by an interface extended by the particular interface. An interface reference for an interface declaration ID associated with a given object OI, wherein the object OI implements methods that are declared by an interface declaration ID includes a pointer to the interface method table corresponding to the interface declaration ID within the class descriptor corresponding to the object OI. An efficient mechanism for method invocations is provided by generating an index that identifies the entry corresponding to the appropriate method in the interface method table pointed to by the pointer of the interface reference, reading the data stored by the entry identified by the index, and controlling the computer processing apparatus to jump to the address identified by the data.

Object model for Java

Alternatives for laying out objects and/or virtual function tables (VFTs), sharing a virtual function table pointer with a direct, virtual base, eliminating the need for an early cast by increasing the number of virtual function table entries, and providing a smart virtual dispatch that eliminates late casts where possible. An object model can be constructed by choosing one option from each of several alternatives. Interaction of alternatives include (1) sharing VFT pointers with a direct, virtual base requires that virtual base pointers are not stored in objects and (2) sharing VFT pointers affects the sharing of virtual base pointers with base classes and placement of late cast adjustment arguments and determining virtual function table entries.

Growing community of inventors

Tarrytown, NY, United States of America

Harini Srinivasan