To build decision tree, information gain is calculated for each
and every attribute and select the attribute with the highest
information gain to designate as a root node. Label the attribute
as a root node and the possible values of the attribute are
represented as arcs.