A high performance identification of jets, produced via fragmentation
of bottom quarks, is crucial for the ATLAS experiment, running at the LHC.
These jets can be identified by exploiting the presence of cascade decay
vertices from bottom hadrons. A general vertex-finding algorithm is
introduced and its application to the search for secondary vertices inside
jets is described. Kinematic properties of the reconstructed vertices are
used to construct several $b$-jet identification algorithms. The features
and performance of the secondary vertex finding algorithm in a jet, as well
as the performance of the jet tagging algorithms, are studied using simulated
$pp \rightarrow t\bar{t}$ events at a centre-of-mass energy of 13 TeV.