An increasing number ofa wide variety of texts on Italian cultural heritage are available today, both online and on paper. However, there are no specific tools (dictionaries, reference materials on technical translations) that can train and support specialists involved in cultural tourism. Mainly focusing on Florence and its cultural heritage, the LBC project (Farina 2016) will try to fill this gap byproviding tools for those who have to write/translate for dissemination in various languages: in a first step by building monolingual corpora (English, French, German, Italian, Russian, Spanish) that the user can freely search; in a second step by developingaplurilingual LSP internet dictionary on cultural heritage which uses the above-mentioned corpora as a primary source. The aim of this paper is to give an insight in the lexicographical process of the LBC-Dictionary, concentrating in particular on data collection and data preparation, which, as is usual for dynamic dictionaries, are open-ended and ever ongoing (Klosa 2013). In particular, we will illustrate the main characteristics of the French and German LBC Corpora and reflect on the provisional French and German entry list, also illustrating the procedure adopted, an alternation of corpus-driven and corpus-based steps (Tognini-Bonelli 2001), for their extraction.