International Business Machines Corporation

Daya Vivek

Stephen Andrew Brodsky

Maryela Evelin Weihrauch

Anshul Dawra

Curt Lee Cotner

Andrei Fedorovich Lurie

Michael Leon Pauser

Sonali Surange

Sumit Bhatia

Anshu Narendra Jain

Rebecca Braun Nin

William Robert Bireley

Christopher John Crone

Azadeh Ahadian

Swaminathan Gounder

Margaret Ann Bernal

Delmar Eugene Blevins

Michael Schenker

Pradeep Teregowda

Vasantha Jayakumar

A computer-implemented method includes determining first passages (FP) that provide evidence for one or more first relations (OOMFR) between first entities in a knowledge graph. The method includes generating an evidence index (EI) that associates the FP with the OOMFR and the first entities, receiving a query subsequent to generating the EI, and identifying, using the EI, the FP responsive to receiving the query. The method includes determining presentation aspects of the FP based on similarity information determined for the FP, and determining that second passages of the FP are substantially similar to at least one other passage of the FP. The method includes pruning the FP to generate a pruned set of passages (PSOP) that includes one or more third passages of the second passages and that does not include any other passages of the second passages. The method includes outputting the PSOP according to the presentation aspects.

Providing relevant evidence or mentions for a query

Embodiments of the invention provide an IDE tool that provides rapid application development features for a variety of disparate data sources. Further, developers may customize and extend the rapid application development features to suit their development needs in a particular case. In a particular embodiment, the rapid application development template may be configured to generate a set of configuration files localized used to deploy a 'lightweight' application, such as a web-service, to a particular application server.

Extensible rapid application development for disparate data sources

Provided are techniques for generating a result set. A statement that includes at least one statement pattern is received, wherein each statement pattern is a template that indicates how a statement is to be modified for execution against a data store and wherein each statement pattern modifies program logic and control flow. Each statement pattern is expanded to generate one or more new statements, wherein each statement pattern is capable of being expanded to zero or more statement patterns. The one or more new statements are executed against a data store to obtain a result set. It is determined whether to modify the result set based on the at least one statement pattern. In response to determining that the result set is to be modified, modifying the result set.

Statement generation using statement patterns

Disclosed is a technique for input parameter binding. When executing a statement, data in an application structure of the statement is compared with optimization information in a bind-in structure. When there is a match between the data in the application structure and data in the optimization information in the bind-in structure, the statement is executed with the optimization information. Also disclosed is a technique for output parameter binding. When executing a statement, data in an application structure of the statement is compared with optimization information in a bind-out structure. When there is a match between the data in the application structure and data in the optimization information in the bind-out structure, the statement is executed with the optimization information.

Method for optimized parameter binding

Growing community of inventors

San Jose, CA, United States of America

Daya Vivek