Paper Info
Title
Applying neural networks to knowledge representation and determination of its meaning
Abstract
Knowledge representation is one of the first challenges AI community was confronted with. To be applicable, knowledge representation techniques must be able not only to represent the knowledge, but also to provide means to determine its meaning. The proposed knowledge representation techniques solve the problem of meaning determination by naming, i.e. by describing the meaning of represented knowledge. These descriptions are provided by database, knowledge base, ontology designers, which give names to tables, fields, classes, properties, relationships, etc. An alternative approach to the problem of determining the meaning would be a neural network approach applied to knowledge representation in a natural language that does not use names, but semantic categories. In this paper we propose a Hierarchical Semantic Form (HSF), a modification of localist approach of connectionist model, which, together with Space of Universal Links (SOUL) algorithm, is capable of representing knowledge in a natural language and interpreting its meaning by using the semantic categories.
Year
DOI
Venue
2007
10.1007/978-3-540-75555-5_50
BVAI
Keywords
Field
DocType
neural network approach,hierarchical semantic form,proposed knowledge representation technique,knowledge representation,localist approach,knowledge base,semantic category,alternative approach,natural language,knowledge representation technique,neural networks,neural network
Procedural knowledge,Commonsense knowledge,Body of knowledge,Knowledge representation and reasoning,Computer science,Knowledge-based systems,Artificial intelligence,Knowledge extraction,Natural language processing,Knowledge base,Open Knowledge Base Connectivity
Conference
Volume
ISSN
ISBN
4729
0302-9743
3-540-75554-3
Citations 
PageRank 
References 
0
0.34
7
Authors
2
Name
Order
Citations
PageRank
Mladen Stanojević1202.72
Sanja Vraneš2606.24