What we examine
Existing descriptions, identifiers, files, controlled vocabularies, and the questions researchers or visitors are trying to answer.
Search becomes more useful when it accounts for the language people use, the unevenness of collection data, and the source record behind every result.
Existing descriptions, identifiers, files, controlled vocabularies, and the questions researchers or visitors are trying to answer.
Text search, semantic retrieval, visual similarity, browse paths, and interfaces that return a result with its source context.
We test retrieval against agreed queries and review the result quality with the people who know the collection.
Each project starts with the data, the context, and the intended use.
[ Discuss your project ]