In general, a lot of different approaches of coin databases and frontends exist, because there are different sources of coin information and also different types of use cases.

We know databases of coins as museal or commercial objects but also databases with historical or archaeological focus, where coins are only one part of the hole database information. Both database types focused on different users and use cases.

I take part in the actual development of two examples of these online coin databases.

The first is the Turnosen-Datenbank (a database of gros tournois). The focus of this project is to present the images of all 3.141 gros tournois of the Erfurt hoard, which was found in 1997. The coins are very uniform, so we create a complex search formula with different ways to describe the coin or coin type. One feature is the use of virtual key boards to create the inscription (or part of the inscription) of the coin. The result of the search is a list of coin images with numismatic determination. In the future we hope to integrate functions of automatical coin recognition.

The second example is the EDV-Fundkatalog Mittelalter/Neuzeit der Numismatischen Kommission (coin finds in Germany of the medieval and modern times). The focus of this project (or the actual stay) is totally different. We have a lot of textual information about these coin finds but no images of these coins. The main focus of this database is to present information about the find localization and the content of hoards. Often we have only small information of finds discovered in the 19^th century. Today the coins of this finds are mostly lost.

Our idea was to use the well known search strategy of library OPACs. The user can start with a simple search form over all database fields. In a second step the user can also create complex search-pattern by using word lists based of database entries.

The results are lists of finds including lists of coins of this finds. In a next step we want to integrate a GIS-modul to create maps of the selected finds. Images of modern finds will also integrate.

 The internal structure of the coin-part of the databases is comparable. There are similar fields of the coin description. Additional features of the second project are informations about the location and the background of the finds. Thats why we can use one data model of the coin information. Different frontends help to build user friendly online-applications.