A multipreference semantics from common sense reasoning to neural network models: An overview

Laura Giordano; Valentina Gliozzi; Daniele Theseider Dupré

A multipreference semantics from common sense reasoning to neural network models: An overview

Department of Science and Technological Innovation

Research output: Contribution to journal › Conference article › peer-review

Abstract

In this short paper we report about a "concept-wise" multipreference semantics for weighted conditionals and its use to provide a logical interpretation to some neural network models, Self-Organising Maps (SOMs) and Multilayer Perceptrons (MLPs). For MLPs, a deep network can be regarded as a conditional knowledge base, in which the synaptic connections correspond to weighted conditionals.

Original language	English
Journal	CEUR Workshop Proceedings
Volume	2969
Publication status	Published - 2021
Event	Joint Ontology Workshops 2021 Episode VII: The Bolzano Summer of Knowledge, JOWO 2021 - Bolzano, Italy Duration: 11 Sept 2021 → 18 Sept 2021

Keywords

Common Sense Reasoning
Neural Network models
Preferential semantics
Typicality in Description Logics

Cite this

@article{fa93be08f0db4a2db23c057422b4b95e,

title = "A multipreference semantics from common sense reasoning to neural network models: An overview",

abstract = "In this short paper we report about a {"}concept-wise{"} multipreference semantics for weighted conditionals and its use to provide a logical interpretation to some neural network models, Self-Organising Maps (SOMs) and Multilayer Perceptrons (MLPs). For MLPs, a deep network can be regarded as a conditional knowledge base, in which the synaptic connections correspond to weighted conditionals.",

keywords = "Common Sense Reasoning, Neural Network models, Preferential semantics, Typicality in Description Logics",

author = "Laura Giordano and Valentina Gliozzi and Dupr{\'e}, \{Daniele Theseider\}",

note = "Publisher Copyright: {\textcopyright} 2021 CEUR-WS. All rights reserved.; Joint Ontology Workshops 2021 Episode VII: The Bolzano Summer of Knowledge, JOWO 2021 ; Conference date: 11-09-2021 Through 18-09-2021",

year = "2021",

language = "English",

volume = "2969",

journal = "CEUR Workshop Proceedings",

issn = "1613-0073",

publisher = "CEUR-WS",

}

TY - JOUR

T1 - A multipreference semantics from common sense reasoning to neural network models

T2 - Joint Ontology Workshops 2021 Episode VII: The Bolzano Summer of Knowledge, JOWO 2021

AU - Giordano, Laura

AU - Gliozzi, Valentina

AU - Dupré, Daniele Theseider

PY - 2021

Y1 - 2021

N2 - In this short paper we report about a "concept-wise" multipreference semantics for weighted conditionals and its use to provide a logical interpretation to some neural network models, Self-Organising Maps (SOMs) and Multilayer Perceptrons (MLPs). For MLPs, a deep network can be regarded as a conditional knowledge base, in which the synaptic connections correspond to weighted conditionals.

AB - In this short paper we report about a "concept-wise" multipreference semantics for weighted conditionals and its use to provide a logical interpretation to some neural network models, Self-Organising Maps (SOMs) and Multilayer Perceptrons (MLPs). For MLPs, a deep network can be regarded as a conditional knowledge base, in which the synaptic connections correspond to weighted conditionals.

KW - Common Sense Reasoning

KW - Neural Network models

KW - Preferential semantics

KW - Typicality in Description Logics

UR - https://www.scopus.com/pages/publications/85117038592

M3 - Conference article

AN - SCOPUS:85117038592

SN - 1613-0073

VL - 2969

JO - CEUR Workshop Proceedings

JF - CEUR Workshop Proceedings

Y2 - 11 September 2021 through 18 September 2021

ER -

A multipreference semantics from common sense reasoning to neural network models: An overview

Abstract

Keywords

Other files and links

Fingerprint

Cite this