The Palace Manager is as an aggregator of distributor content. It uses the metadata describing the content communicated by the distributor of the collection from their API or OPDS / ODL feed. This can be enriched through metadata services as well.
Challenges
Unfortunately, there is not consistency in the scope and quality of descriptive metadata between e-content providers nor availability via their API.
There is no consistent use of classification schemes and vocabularies used by providers either.
The OPDS Catalog specification does not provide some minimum standard or specific requirement by which a classification schema or vocabulary is provided via a feed. However, it does not preclude some form of category or subject metadata being published as part of the feed.
The Solution
The Palace Manager can accommodate this through custom vocabularies (Folksonomies) or standard vocabularies (FAST, LCSH, LCC, BISAC, BIC). It can also can apply search tags to capture a vendors description of the content as well for use in classification determination and search. Using this information, the Palace Manager boils the information into its own folksonomy used by The New York Public Library.
The Goal
The Palace Manager classification logic and process routines attempt to normalize the various metadata inputs provided by content providers and collections. These external classifications are used to take a book’s subject, audience, genre, fiction or non-fiction, and language metadata and map it to an internal understanding of a books Genre, Fiction Status, Age, Language and Audience.
Code Repository: