Read, describe and share! Building an interactive literary web service: an article about Kirjasampo.

Purpose - This paper aims to describe the development of a fiction web service which will serve as a tool for information searches, and a place for saving and sharing tacit knowledge of and experiences with fictional works. Key partners are the Turku City Library, the Semantic Computing Research Group, Libraries.fi editorial staff and Vaasa City Library. Design/methodology/approach - The system's technical infrastructure was built in the autumn of 2010: database, data storage platform and content description. Base data have been converted, data-saving protocols created and system data verified in order to stay up to date. Information on the project has been widely distributed to libraries, book industry actors and other potential partners, even at the international level. Design of the user interface has begun, and with its construction it will become apparent how well the project targets are being met, i.e. to make available a user-friendly service that offers library fiction collections and their attendant expertise and inspires service users to read, make recommendations and evaluate fiction by the spring of 2011. Findings - Fiction is an integral part of Finnish library collections. However, its information search and help desk resources are lacking, and the content of older works of fiction has not been systematically described. Where fiction customer service is concerned, the librarian is largely left to rely on their own reading, memory and associations. This Kirjasampo Project will provide better access to fiction collections. Content description is done using ontologies, thus requiring the ontologisation of Kaunokki, the Finnish fiction thesaurus, as part of the project. In addition to Kaunokki, time, place, language and nationality ontologies are used. Originality/value - In the wide-ranging and multidisciplinary Finnish Kirjasampo Project, thought has been given to how these problems can be solved using modern technologies. Because the service will have a large number of features that cannot be realised with existing library systems, it will be done using semantic web tools. A fiction metadata schema, which defines the necessary fields in information models, was developed for the service. The Functional Requirements for Bibliographic Records (FRBR) was applied in the information models, with the content of each work being described ("abstract work") and the various versions and institutions ("physical works") linked to it. This information is saved in the RDF database using a Kirjasampo-SAHA annotation editor, which is a browser-based and user-friendly metadata-saving tool. [ABSTRACT FROM AUTHOR]