@inproceedings{12b4402ed73c4589ab2cd8b52e37a31b,
title = "Graph decomposition approaches for terminology graphs",
abstract = "We propose a graph-based decomposition methodology of a network of document features represented by a terminology graph. The graph is automatically extracted from raw data based on Natural Language Processing techniques implemented in the TermWatch system. These graphs are Small Worlds. Based on clique minimal separators and the associated graph of atoms: a subgraph without clique separator, we show that the terminology graph can be divided into a central kernel which is a single atom and a periphery made of small atoms. Moreover, the central kernel can be separated based on small optimal minimal separators.",
keywords = "Graph algorithms, Graph decomposition, Polyhedral approach, Text mining, Topic vizualisation",
author = "Biha, {Mohamed Didi} and Bangaly Kaba and Meurs, {Marie Jean} and Eric Sanjuan",
year = "2007",
language = "English",
isbn = "9783540766308",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
pages = "883--893",
booktitle = "MICAI 2007",
note = "6th Mexican International Conference on Artificial Intelligence, MICAI 2007 ; Conference date: 04-11-2007 Through 10-11-2007",
}