The Matrix Element Method (MEM) is an advanced analysis technique that optimally exploits the underlying dynamics of possible physics processes responsible of a given final state. Applied at the Tevatron for a precise top quark mass measurement, its usage at the Large Hadron Collider (LHC) is more challenging. The complexity is due to the modeling of all interactions responsible for a given observed final state together with the phase-space integration considering properly the detector response parametrization. This thesis shows for the first time that the MEM can effectively be used at the LHC for the final state involving pairs of high energetic leptons and b-quarks (llbb), with or without significant missing transverse energy. For the final state with missing transverse energy, it has been shown that a precise top quark mass measurement could be achieved by means of the MEM, for events induced by a top quark pair production followed by their leptonic decay. An original approach for performing a model independent search of heavy resonances decaying into a top quark pair is also addressed by using a variant of the usual MEM. Final state without a large missing transverse energy can be populated by the associated production of a Z and the Standard Model Higgs bosons (ZH). A dedicated search for a SM Higgs boson based on the MEM, used to distinguish between signal and background processes, has been successfully performed, using 5/fb data recorded by the CMS experiment during the 2011 data taking period. Assuming a Higgs boson mass of 125 GeV/c2, the observed upper limit on the ZH production cross section times the H->bb branching ratio is 5.7 times the Standard Model prediction.