Browse Prior Art Database

Method for Dynamic Construction of Natural Language Seed Question or Summary based on Profile Context

IP.com Disclosure Number: IPCOM000237567D
Publication Date: 2014-Jun-24
Document File: 4 page(s) / 172K

Publishing Venue

The IP.com Prior Art Database

Abstract

A method for dynamic construction of natural language seed question or summary based on profile context is disclosed.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 53% of the total text.

Page 01 of 4

Method for Dynamic Construction of Natural Language Seed Question or Summary based on Profile Context

Disclosed is a method for dynamic construction of natural language seed question or summary based on profile context.

Customers when facing problems with the products and services they are using turn


1.

to their respective service providers for help. They expect the service provider to resolve their issue(s) is a quick manner.

With the IBM (Q&A) approach, provided the right natural language question is entered by the end user (customer or an agent), Watson will perform traditional "Question Analysis, Hypotheses Generation, Hypotheses Scoring, and Supporting Evidence Merge and Ranking"

The onus is on the end user to ask the right question that would lead to their problem / issue with the products/services getting fixed.

While Watson does provide accurate answers provided the end users ask the right questions, it currently does not the understand the user profile at a given point in time. It does not understand the following:

Who the user is


Snapshot of their profile (name, address, account summary etc) What problems they are facing with their products/services
Location from where they are calling
Past interactions with the service provider outside of Watson channel Why they are calling

Disclosed is a method of using a "customer profile context"to auto construct a (or a set of) Natural Language Question the user might ask and pose it to the user as a seed question. The "customer profile context" is bootstrapped with information very frequently / periodically from applications like customer information management, billing, incident management, trouble ticketing etc) and AUTO CONSTRUCT a Natural Language Question the user might ask and pose it to the user as a seed question,
This may be done prior to or even b...