Paper Info
Title
Seven Hard Problems in Symbolic Background Knowledge Acquisition
Abstract
By using a special characterization of machine learning algorithms, we first define what is background knowledge, as opposed to case-based, strategic, and explanatory types of knowledge. We oppose also the symbolic to the numeric view of background knowledge. We discuss then what we see as the seven most difficult topics in background knowledge acquisition, namely the detection of implicit implications, first order logic knowledge representation and acquiring "Skolem" functions, uncertain knowledge, weak knowledge, time management and fusion of several sources of knowledge, knowledge for vision, certification of knowledge.
Year
DOI
Venue
1990
10.1007/3-540-53414-8_29
IMYCS
Keywords
Field
DocType
symbolic background knowledge acquisition,hard problems,first order logic,time management,machine learning
The Symbolic,Knowledge representation and reasoning,Descriptive knowledge,Computer science,First-order logic,Time management,Artificial intelligence,Natural language processing,Certification,Knowledge acquisition
Conference
ISBN
Citations 
PageRank 
3-540-53414-8
0
0.34
References 
Authors
10
1
Name
Order
Citations
PageRank
Yves Kodratoff1581172.25