Abstract

The LRM representative expression is an attribute of an expression that is transferred to the work, and it serves as a means of identifying the work. However, without clear criteria for attribute selection, catalogers must use their judgment, making it difficult to recognize attributes and values of the original or canonical expression. This study interviewed cataloging experts to examine their perceptions of representative expression attributes, focusing on ambiguous cases like works with multiple original expressions or no original expression. The findings suggest that these attributes can support efficient user searching and reduce cataloging workload. In most cases, the attributes of “content type” and “date of expression” were adopted, while different attributes were chosen depending on the form of the work, such as “language” for text and “performing medium” for music. Key issues identified during interviews included considerations for both cataloging and attribute display. To address these, the study proposes building a “core attribute set” and “displaying tailored attributes” based on the primary user group.