The semantic enrichment of cultural heritage data is a crucial step for memory institutions to provide enhanced search services and to reduce the costs for managing resources. Yet, using information systems based on semantic web technologies implies to migrate and enrich the legacy data according to the dedicated semantic models. In the library community such task can be very complex when handling large catalogs of records stored using old and flat models. This paper focuses on the challenges related to metadata migration process and semantic enrichment of bibliographic data. We present the contributions and lessons learned from the FRBRization study and we provide preliminary discussions and ongoing work about the enrichment of FRBRized entities.