Extracting Commonsense Knowledge Using Concepts Properties

Extracting Commonsense Knowledge Using Concepts Properties

Eduardo Blanco (The University of Texas at Dallas, USA), Hakki C. Cankaya (Izmir University of Economics, Turkey) and Dan Moldovan (The University of Texas at Dallas, USA)
DOI: 10.4018/978-1-61350-447-5.ch022
OnDemand PDF Download:
$30.00
List Price: $37.50

Abstract

Commonsense knowledge encompasses facts that people know but do not communicate most of the time. For example, one needs water and soap to take a shower is commonsense. This chapter presents a semantically grounded method for extracting commonsense knowledge. First, commonsense rules are identified, e.g., one cannot see imaginary objects. Second, those rules are combined with a basic semantic representation in order to infer commonsense facts, e.g. one cannot see a flying carpet. Further combinations of semantic relations with inferred commonsense facts are proposed and analyzed. Experimental results show that this novel method is able to extract thousands of commonsense facts with little human interaction and high accuracy.
Chapter Preview
Top

Background

Commonsense knowledge is considered obvious and many researchers (Singh, 2002; Lenat, 1995; Ahn, Kedia, & Blum, 2006) claim that it is unfeasible to obtain it from text or any other existing resource. They believe that humans, experts or non-experts, are needed to obtain commonsense facts.

Cyc is the biggest and oldest project aiming at building a commonsense knowledge base (Lenat, 1995). The project started in 1984 and since then experts have introduced millions of commonsense facts using a formal language, CycL. Currently, the Cyc knowledge base contains nearly 500,000 terms, including about 15,000 types of relations, and about 5,000,000 facts (assertions) relating these terms (Source: http://www.cyc.com/).

Complete Chapter List

Search this Book:
Reset