Browse Prior Art Database

Prompting users to refine question to improve answer accuracy in a cognitive system

IP.com Disclosure Number: IPCOM000238579D
Publication Date: 2014-Sep-04
Document File: 2 page(s) / 48K

Publishing Venue

The IP.com Prior Art Database

Abstract

A method for prompting users to refine question to improve answer accuracy in a cognitive system is disclosed.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 68% of the total text.

Page 01 of 2

Prompting users to refine question to improve answer accuracy in a cognitive system

Disclosed is a method for prompting users to refine question to improve answer accuracy in a cognitive system.

In the case of conversational questions, the context is not always known by the recipient - in this case a cognitive system. This method aims to clarify ambiguities in a question
in order to submit a well formed and thorough question through the submission pipeline. A "preflight" system is disclosed that analyzes the question about to be asked and returns a set of clarification answers, assumptions, synonyms, date ranges for temporal accuracy as the question is typed in or after a short delay in input.

In cognitive systems such as IBM Watson

                          ® , the wording of a question has a large impact on the accuracy with which the system can find an answer. The disclosed method reviews a question as the user is typing it in and presents choices which guide the user to be more precise in how the question is being asked. For example, if the user asks "What storms caused the most damage in the US." The system may prompt the user to select the type(s) of storms, tell the user the system is assuming that US means United States and prompt the user to supply a date range. By making these selections, the question will be reworded to be more precise and better answers will be returned.

This approach also has training benefits. IBM Watson shows the user how to create a better formed and more...