Extract specific portions which are inside the comments (Ex: voice over) and classed by voice actor and references.