Browse Prior Art Database

Method and System for Providing a Cognitive Integration Service

IP.com Disclosure Number: IPCOM000250382D
Publication Date: 2017-Jul-07
Document File: 4 page(s) / 227K

Publishing Venue

The IP.com Prior Art Database

Abstract

A method and system is disclosed for providing a cognitive integration service. The cognitive integration service provides a trainable and reusable aid to complex cognitive cloud service adoption.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 40% of the total text.

1

Method and System for Providing a Cognitive Integration Service

Cognitive services or cognitive cloud services require a large and complex operational dataset to describe a specific nature of a problem. As the cognitive services mature via training, self-learning and product feature development, the operational dataset expands and results in a complex integration. The complex integration involves in mapping of hundreds of attributes from one or more existing solutions, developing a custom interface aggregating using the mappings and transforming each attribute into a format required by a cognitive service application programming interface (API) specification. Thus, there is a need for a cognitive integration service that can be trained for complex cognitive services and then extended to support multiple input files and formats.

Disclosed is a method and system for providing a cognitive integration service. The cognitive integration service provides a trainable and reusable aid to complex cognitive cloud service adoption.

The method and system enables a creation of a cognitive integration service that reduces development and maintenance of resource requirements to integrate with complex cognitive services. The cognitive integration service is trainable and operates as a facade to the complex cognitive services.

Once the cognitive integration service is trained, the facade fetches data from an operational dataset, associates data elements with data arguments required by the complex cognitive services and performs transformation/translation of the data elements associated with the data arguments. The transformation is performed from/to (via class library) creating a data package as per a complex cognitive service integration specification.

The transformation of the data elements includes a substitution technique that produces a high degree of accuracy on structured data. The substitution technique is basic machine transaction technique. Also other techniques, such as, but not limited to, natural language processing techniques and emerging domain specific natural language machine translation are used for the transformation.

After transformation, the cognitive integration service applies a cartridge pattern by incorporating training on industry standard data specifications and custom data specifications. Thus, applying the cartridge pattern produces a highly reusability implementation for use with datasets compliant with the industry standard data specifications and a slightly reduced reusability implementation for use with lesser compliant datasets.

FIG. 1 illustrates a cognitive integration service created for a complex cognitive service.

2

FIG. 1

As illustrated in FIG. 1, the cognitive integration service is created for the complex cognitive service that requires complex operational data integration requirements. The cognitive integration service operates as a WO integration facade. WO integration facade receives a cognitive pattern (for exam...