After lunch, the participants of the IMPACT Project Outcomes Event split up into two groups. One went upstairs to the tool workshop, while the other group stayed in the Auditorium to see the library experiences of the project.
Library experiences in IMPACT: KB National library of the Netherlands by Caroline van Wijk and Judith Rog (KB)
The afternoon was kicked off by the Koninklijke Bibliotheek with Caroline van Wijk (project leader Innovative Projects) who presented a pilot done with the CONCERT tool in 2011.
Judith Rog (project leader Innovative Projects) followed with the pilot of post-correction and OCR tools done in 2012. Both had good experiences with the tools and the project, but the outcomes have not yet been picked up by the KB due to the current infrastructure of the library.
Library experiences in IMPACT: National and University Library of Slovenia by Alenka Kavčič-Čolić (NUK) – Audio File
The National and University Library of Slovenia had less trouble implementing outcomes of the project, as we heard from Alenka Kavčič-Čolić (Head of the Library Research Centre and Head of the Research and Development Division). NUK joined the IMPACT project together with the Jožef Stefan Institute in the extension in 2010. One of their joint achievements is the development of a historical lexicon which is currently implemented in the search engine of dLib.si, making it possible to search for a modern word and finding historical variants. Next to this, working in an international research project provided many other benefits for both NUK and JSI.
The Functional Extension Parser (FEP) and Ebooks On Demand (EOD) by Andreas Parschalk (University of Innsbruck)
Another one of IMPACT”™s success stories is the implementation of UIBK”™s Functional Extension Parser. Andreas Parschalk (Lead Software developer) introduced the EOD-network and the workings of such a setup. Currently, more than 30 libraries in 12 countries participate in the digitisation on demand-network. Implementing the Document Understanding Platform into this network would thus provide a new service to a wide range of libraries and users. Ten libraries have signed up for the beta-testing, but many more showed interest, so keep an eye out for the EOD button in your local catalogue!
To help catalogue the many dissertations the DNB gets sent on a daily basis, Christa Schöning-Walter, (Project leader PETRUS) explained the tests they”™ve done with the Functional Extension Parser to automatically extract information from scanned title pages. Using specific rules, the software extracts various types of metadata, such as title, author, and university name. A test done with 1000 publications provided some very interesting results and the DNB will continue this work with a feasibility study. New ideas have already been formed for more uses of the tool.
Besides libraries, also other parties have taken an interest in the tools of IMPACT, such as the Europeana Newspaper project, as Lotte Wilms (KB – Project leader European Projects) showed in her presentation. 18 Million pages of European newspapers will be made available in Europeana during the course of the project, and many of those will be accompanied with the OCRed full text via a search portal that will be developed by The European Library. The 17 partners in the project will work together in the EU funded project to form this new collection until 2015, with the help of some of the IMPACT tools, among which the improved ABBYY OCR software and the evaluation tools of the University of Salford. More information is available from www.europeana-newspapers.eu