Difference between revisions of "Publications/kamal.22.xkdd"

From LRDE
 
(One intermediate revision by the same user not shown)
Line 3: Line 3:
 
| date = 2022-09-12
  
| date = 2022-09-12
 
| authors = Ataollah Kamal, Elouan Vincent, Marc Plantevit, Céline Robardet
  
| authors = Ataollah Kamal, Elouan Vincent, Marc Plantevit, Céline Robardet
−
| booktitle = ECML PKDD International Workshop on eXplainable Knowledge Discovery in Data Mining
 +
| booktitle = Workshop on eXplainable Knowledge Discovery in Data Mining. Machine Learning and Principles and Practice of Knowledge Discovery in Databases - International Workshops of ECML PKDD 2022, GrenobleFrance, September 19-23, 2022, Proceedings, Part I
−
| title = Improving the quality of rule-based GNN explanations
 +
| title = Improving the Quality of Rule-Based GNN Explanations
 
| address = Grenoble, France
  
| address = Grenoble, France
 
| abstract = Recent works have proposed to explain GNNs using activation rules. Activation rules allow to capture specific configurations in the embedding space of a given layer that is discriminant for the GNN decision. These rules also catch hidden features of input graphs. This requires to associate these rules to representative graphs. In this paper, we propose on the one hand an analysis of heuristic-based algorithms to extract the activation rulesand on the other hand the use of transport-based optimal graph distances to associate each rule with the most specific graph that triggers them.
  
| abstract = Recent works have proposed to explain GNNs using activation rules. Activation rules allow to capture specific configurations in the embedding space of a given layer that is discriminant for the GNN decision. These rules also catch hidden features of input graphs. This requires to associate these rules to representative graphs. In this paper, we propose on the one hand an analysis of heuristic-based algorithms to extract the activation rulesand on the other hand the use of transport-based optimal graph distances to associate each rule with the most specific graph that triggers them.
−
| pages = 1 to 16
 +
| pages = 467 to 482
 
| lrdekeywords = IA
  
| lrdekeywords = IA
 
| lrdenewsdate = 2022-09-12
  
| lrdenewsdate = 2022-09-12
Line 13: Line 13:
 
| type = inproceedings
  
| type = inproceedings
 
| id = kamal.22.xkdd
  
| id = kamal.22.xkdd
  +
| identifier = doi:10.1007/978-3-031-23618-1_31
 
| bibtex =
 
| bibtex =
 
@InProceedings<nowiki>{</nowiki> kamal.22.xkdd,
  
@InProceedings<nowiki>{</nowiki> kamal.22.xkdd,
 
author = <nowiki>{</nowiki>Ataollah Kamal and Elouan Vincent and Marc Plantevit and
  
author = <nowiki>{</nowiki>Ataollah Kamal and Elouan Vincent and Marc Plantevit and
 
C\'<nowiki>{</nowiki>e<nowiki>}</nowiki>line Robardet<nowiki>}</nowiki>,
  
C\'<nowiki>{</nowiki>e<nowiki>}</nowiki>line Robardet<nowiki>}</nowiki>,
−
booktitle = <nowiki>{</nowiki>ECML PKDD International Workshop on eXplainable Knowledge
 +
booktitle = <nowiki>{</nowiki>Workshop on eXplainable Knowledge Discovery in Data
  +
Mining. Machine Learning and Principles and Practice of
−
Discovery in Data Mining<nowiki>}</nowiki>,
  
  +
Knowledge Discovery in Databases - International Workshops
⚫
title = <nowiki>{</nowiki>Improving the quality of rule-based GNN explanations<nowiki>}</nowiki>,
  
  +
of <nowiki>{</nowiki>ECML<nowiki>}</nowiki> <nowiki>{</nowiki>PKDD<nowiki>}</nowiki> 2022, Grenoble, France, September 19-23,
  +
2022, Proceedings, Part <nowiki>{</nowiki>I<nowiki>}</nowiki><nowiki>}</nowiki>,
  ⚫
title = <nowiki>{</nowiki>Improving the Quality of Rule-Based <nowiki>{</nowiki>GNN<nowiki>}</nowiki> Explanations<nowiki>}</nowiki>,
 
year = <nowiki>{</nowiki>2022<nowiki>}</nowiki>,
  
year = <nowiki>{</nowiki>2022<nowiki>}</nowiki>,
 
address = <nowiki>{</nowiki>Grenoble, France<nowiki>}</nowiki>,
  
address = <nowiki>{</nowiki>Grenoble, France<nowiki>}</nowiki>,
Line 34: Line 38:
 
graph distances to associate each rule with the most
  
graph distances to associate each rule with the most
 
specific graph that triggers them.<nowiki>}</nowiki>,
  
specific graph that triggers them.<nowiki>}</nowiki>,
−
pages = <nowiki>{</nowiki>1-16<nowiki>}</nowiki>,
 +
pages = <nowiki>{</nowiki>467--482<nowiki>}</nowiki>,
  +
doi = <nowiki>{</nowiki>10.1007/978-3-031-23618-1\_31<nowiki>}</nowiki>,
 
note = <nowiki>{</nowiki>accepted<nowiki>}</nowiki>
  
note = <nowiki>{</nowiki>accepted<nowiki>}</nowiki>
 
<nowiki>}</nowiki>
  
<nowiki>}</nowiki>

Latest revision as of 06:53, 4 April 2023

Authors
Ataollah Kamal, Elouan Vincent, Marc Plantevit, Céline Robardet
Where
Workshop on eXplainable Knowledge Discovery in Data Mining. Machine Learning and Principles and Practice of Knowledge Discovery in Databases - International Workshops of ECML PKDD 2022, GrenobleFrance, September 19-23, 2022, Proceedings, Part I
Place
Grenoble, France
Type
inproceedings
Keywords
IA
Date
2022-09-12

Abstract

Recent works have proposed to explain GNNs using activation rules. Activation rules allow to capture specific configurations in the embedding space of a given layer that is discriminant for the GNN decision. These rules also catch hidden features of input graphs. This requires to associate these rules to representative graphs. In this paper, we propose on the one hand an analysis of heuristic-based algorithms to extract the activation rulesand on the other hand the use of transport-based optimal graph distances to associate each rule with the most specific graph that triggers them.


Bibtex (lrde.bib)

@InProceedings{	  kamal.22.xkdd,
  author	= {Ataollah Kamal and Elouan Vincent and Marc Plantevit and
		  C\'{e}line Robardet},
  booktitle	= {Workshop on eXplainable Knowledge Discovery in Data
		  Mining. Machine Learning and Principles and Practice of
		  Knowledge Discovery in Databases - International Workshops
		  of {ECML} {PKDD} 2022, Grenoble, France, September 19-23,
		  2022, Proceedings, Part {I}},
  title		= {Improving the Quality of Rule-Based {GNN} Explanations},
  year		= {2022},
  address	= {Grenoble, France},
  month		= sep,
  abstract	= {Recent works have proposed to explain GNNs using
		  activation rules. Activation rules allow to capture
		  specific configurations in the embedding space of a given
		  layer that is discriminant for the GNN decision. These
		  rules also catch hidden features of input graphs. This
		  requires to associate these rules to representative graphs.
		  In this paper, we propose on the one hand an analysis of
		  heuristic-based algorithms to extract the activation rules,
		  and on the other hand the use of transport-based optimal
		  graph distances to associate each rule with the most
		  specific graph that triggers them.},
  pages		= {467--482},
  doi		= {10.1007/978-3-031-23618-1\_31},
  note		= {accepted}
}





Retrieved from "https://www.lrde.epita.fr/index.php?title=Publications/kamal.22.xkdd&oldid=128542"