Performance Measurement of b tagging Algorithms Using Data containing Muons within Jets

Three different methods to measure the performance of $b$-quark identification methods are discussed. These methods can be applied to collider data with reconstructed muons within jets. Two of these methods rely on fits to the \ptrel distribution of the muon; the third is less dependent on the Monte Carlo simulation. The methods are applied to two $b$-tagging algorithms that relay on tracks with large impact parameters with respect to the interaction point. Overall, the measured efficiencies show rather good agreement with the efficiencies obtained from the MC truth information, in particular for the region of lower values of jet $p_T$. A luminosity projection of the expected uncertainties is included.