The problem of finding a graceful labelling of a graph, or proving that the graph is not graceful, has previously been modelled as a CSP. A new and much faster CSP model of the problem has been recently proposed. This model was inspired by a mathematical proof. We present a third model that is in some sense a combination of the previous two, but is much more efficient than either. We give several new results for graphs whose gracefulness was previously unknown. The possibility to generalize the use of the modelling techniques presented here is discussed.
Back to "Barbara Smith - Publications on Constraint Programming"