Classification and subject indexing systems have long been the mainstay of established information providers to deliver content precisely on topic. Logical semantic hierarchies and rich interconnections of related terms and synonyms enable accurate retrieval and browsing of similar resources and ideally should be available in online environments. But the cost of features may not be sustainable with massively growing resources. Efforts to merge databases and map disparate subject terminology require considerable human intervention. A possible solution combines controlled and uncontrolled terms from three sources: authoritative professional indexing, automated term suggestion and uncontrolled keywords proposed by authors or end users’ social tags. Research is required to investigate the effectiveness, cost and applicability of combining controlled and uncontrolled terms for information retrieval.