Image annotation becomes increasingly more important as the Web continues to grow. We propose a novel approach to enhancing keyword-based web-image annotation in folksonomy, where a volunteer user is notified what kind(s) of keywords are necessary, and that keywords have been sufficiently provided by other volunteer users. Necessary information is obtained from the keyword-filled facet tables, and sufficient information provided by the quantified conceptual keyword relations. Though the facet tables and the conceptual keyword relations are set by referring to an ontology, users do not need to know its details to successfully annotate the images.