Abstract

This work addresses the prototypical problem of a cultural heritage institution with the ambition to disclose all of its content in a single, unified system. Like enterprises, these institutions often have heterogeneous collections distributed over multiple legacy systems. The brute-force approach taken here involves a mostly unconditional merging of the heterogeneous sub-collections and flattening of all metadata structures, effectively turning the problem to free-text retrieval. Our main findings are as follows: First, by converting all digital content from several systems of one cultural heritage institution to text, and indexing it with a standard IR system, we show that a unified approach is a viable option to give access to heterogeneous collections. Second, although our approach is simplistic, the initial empirical evaluation validates its superior performance against the legacy fragmented systems currently in use by the institute. Third, in a user study, with test persons ranging from expert users (such as internal employees) to naive users (such as potential visitors of the institution’s web-site), we find that all test person’s preferred the unified system—even those who work with the existing propriety system on a day-to-day basis.