• DocumentCode
    265938
  • Title

    Computation of all possible maximal cliques of a weakly triangulated graph in polynomial time

  • Author

    Bandyopadhyay, Supriyo ; Kumar Pal, Rajat

  • Author_Institution
    Dept. of Comput. Sci. & Eng., Univ. of Calcutta, Kolkata, India
  • fYear
    2014
  • fDate
    27-29 Aug. 2014
  • Firstpage
    159
  • Lastpage
    168
  • Abstract
    In this paper, we have addressed the computation of an invariant of weakly triangulated graph. The invariant computed here are all possible maximal cliques of the specific graph class. The logic of computing all possible maximal cliques of weakly triangulated graph is based on one of its prime characterization, i.e., existence of more than one 2-pair in a weakly triangulated graph. For each graph belonging to such a class there exists a sequence of 2-pairs whose successive merging leads the graph into a complete one. In a reverse way successive decomposition of vertices of the obtained complete graph ultimately yields some non-decomposable induced subgraph of the original weakly triangulated graph. These non-decomposable induced subgraphs are not only maximal cliques, but subcliques or redundant occurrences of the same maximal cliques. Our algorithm proposed in this paper, while breaking by 2-pairs produces only maximal cliques in O(n3 log |E|) time for a weakly triangulated graph G = (V, E), where |V| = n. The paper along with the algorithm contains necessary theorems, lemmas, etc., proving that the algorithm invented here terminates successfully and produces the desired maximal cliques only.
  • Keywords
    graph theory; polynomials; nondecomposable induced subgraph; polynomial time; possible maximal cliques; prime characterization; specific graph class; weakly triangulated graph; Algorithm design and analysis; Binary trees; Joining processes; Merging; NP-complete problem; Polynomials; Vegetation; 2-pair; NP-complete problem; height reduction technique; maximal cliques; weakly triangulated graph;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Science and Information Conference (SAI), 2014
  • Conference_Location
    London
  • Print_ISBN
    978-0-9893-1933-1
  • Type

    conf

  • DOI
    10.1109/SAI.2014.6918185
  • Filename
    6918185