Dismiss
InnovationQ will be updated on Sunday, Oct. 22, from 10am ET - noon. You may experience brief service interruptions during that time.
Browse Prior Art Database

Method and System to Enhance Domain Knowledge from Historical Data

IP.com Disclosure Number: IPCOM000236591D
Publication Date: 2014-May-05
Document File: 4 page(s) / 108K

Publishing Venue

The IP.com Prior Art Database

Abstract

A Method and System to Enhance Domain Knowledge from Historical Data 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 and System to Enhance Domain Knowledge from Historical Data

Disclosed is a method and system to enhance domain knowledge from historical data.

Domain knowledge background

Domain knowledge captured in a re-usable form is necessary in many applications. For example, clustering document collection, information retrieval, question answering, document classification, and support call routing all require domain knowledge for proper handling.

Domain knowledge can be in various forms. Examples include keywords and phrases (entities) and relationship between entities. Relationship in turn could be of various types, such as, causal, temporal, kind of, etc. Even the possible list of relationships depends on the domain.

Domain knowledge evolves over time.

Typically. domain knowledge is manually or semi-automatically built.

Domain knowledge could be noisy irrespective of how it is generated. If it is manually generated, the noise may creep in because of limitations on the knowledge of the person who created the knowledge. If semi-automatic or automatic, this could be because of noise in the raw data or errors in preprocessing steps.

The extent of noise that can be tolerated depends on the end application that is using the domain knowledge

Drawback of Existing Solutions

The relationship should be present in some sentences in the text corpora. Existing solutions are not able to detect or discover relations that do not explicitly exists in sentences.

They can discover only pre-determined types...