Browse Prior Art Database

Cognitive Service for Content Entry Quality Improvement Disclosure Number: IPCOM000247365D
Publication Date: 2016-Aug-29
Document File: 4 page(s) / 207K

Publishing Venue

The Prior Art Database


A method for cognitive service for content entry quality improvement (CEQI) is disclosed.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 51% of the total text.

Page 01 of 4

Cognitive Service for Content Entry Quality Improvement

Disclosed is a method for cognitive service for content entry quality improvement (CEQI). The method may support training on one or more domains to enable conversational data collection and evaluation of information. Historically, data entry has been accomplished through field based user interfaces (UI) with edit rules implemented

at a field level, cross-field, UI and cross-UI level. The usage of the CEQI service replaces the need for field data entry with unstructured data entry against which the service detects and reports upon the the terms and phrases within unstructured text. The usage of this service allows the caller to encourage improved quality and completeness of data entered by a user. Strategies include dialog, gamification, or others means by leveraging the service response of identified "entities" existing or missing from the unstructured text the service is asked to evaluate. The service may also be used to evaluated and report upon the presents of "entities" within the existing unstructured text.

An implementation of a CEQI cognitive service may implement one or more models

trained to identify entities in one or more domains into a run time environment. The runtime exposes the model(s) as a service. Data collection is performed by a UI which first calls a service to retrieve domain specific metadata including entities associated

with the domain, dictionaries associated with the entity, and text examples for each entity. The entities may be used to populate the UI with a checklist of information to be included in the content. The checklist informs the user of the desired content to be entered. The dictionaries may be used to support type-ahead features and the example text is used to provide examples for each entity illustrating the format of the specific text quality desired. As the content is collected, at an event, such as a pause in user input or tab to next feed, the implementing solution calls a second service to score the content entered using the trained model. A response of entities found and/or not found is prepared and returned to the implementing solution which via the UI feeds back relative quality assessment of the content, for example by colorizing a list of entities green or red to indicate what is found and which is still missing, encouraging quality improvement. An iterative approach continues to encourage completeness of

the content.

Examples of solutions implementing the service may include:

Content Entry Applications (CEA), used to collect data from a user within the domain for which a CEQI service is supports and used to encourage entry of relevant data.

Content Analysis Applications, used to evaluate existing content for the existence of specific data within that domain.

Other Tooling, services with provide content validation services.

Figure 1 illustrate an example design for a CEQI service implementation. A macro and

micro service desi...