Fact-based databases

BI = bibliographic database CA = catalogue of chemicals db = database DADB = metadatabase of online databases DACD = metadatabase of CD-ROMs DAIN = metadatabase of internet resources FBDB = fact-based database FT = full-text database INDB = integrated database MD = metadatabase NU = numeric database RD = research database RE = reaction database ST = structural database TBDB = text-based database WWW = World Wide Web. 

The fact-based databases are divided into numerical (factual) databases, metadatabases, research databases, and catalogues of chemical substances. These types will be explained in the following sections,... 

Metadatabases (MD). Metadatabases are also categorized as fact-based databases. In this type of database, objects - in this case databases - are described with the aid of datasets. A prominent example is the already mentioned metadatabase Gale Directory of Databases which is available in printed form, online, and on CD-ROM. In addition, every host has its own metadatabase with a description of each... 

Fact-based databases have a percentage of over 30% both in DADB and DACD, which means that approximately one third of all online databases and CD-ROMs in the field of environmental chemistry are fact-based. 

Concerning text-based and fact-based databases, the Metadatabase of Internet Resources gives completely different results from DADB and DACD. In DAIN most of the databases are fact-based (89%) and only 8% are text-based. This situation is in extreme contrast to the other two media. For this reason the DAIN results were investigated further. 

TBDB = text-based databases FBDB = fact-based databases INDB = integrated databases... 

As mentioned in Section 1.2 describing the various Molecular Biology resources available, applications have played a large part in developing these resources. In fact some databases are entirely based on the results of applications. 

The overwhelming richness of structural types can be seen from the fact that there are over 140 recognized zeolite networks (thermally stable and microporous), this number growing by about six every year [2,3]. On the theoretical side there are estimates that the number of plausible regular tetrahedral frameworks exceeds 100 000, f2j and the uninodal nets alone number more than 150 [4]. The lUPAC approved Structure Commission of the International Zeolite Association are in charge of the assignment of three letter codes to each unique zeolite structure and publish them in a web-based database [5J. 

The text-based databases are divided into bibliographic and full-text databases. They still completely dominate the fact-based and integrated databases in the online and CD-ROM categories. ... 

In Figure 3, the percentage is shown for the three discussed types of databases text-based, fact-based, and integrated databases. As the metadatabases are quite different in size, especially the newest Metadatabase of Internet Resources - DAIN, which still comprises under 100 entries, comparisons are quite vague and only show trends. 

In Figure 3, text-based (TBDB), fact-based (FBDB), and integrated databases (INDB) are compared in the three discussed metadatabases. The percentage of each category is shown. The following conclusions can be drawn ... 

Text-based databases are still predominant in DADB and DACD. That means that in online databases and CD-ROMs still more bibliographic and full-text databases are found than fact-based and integrated databases. In both DADB and DACD over 60% of the evaluated databases are text-based. 

The discussed situation concerning fact-based and text-based resources differs greatly from that of DADB and DACD. We must therefore consider types of databases which are covered in DAIN. In Figure 4 the different types given in Section 2.4 are shown. 

Apart from the fact that only very few bibliographic (BI) and full-text (FT) databases are covered in DAIN, the situation in the fact-based area is of significance. Here 45% are numeric databases (NU) and 35% are metadatabases (MD). Chemical catalogues (CA) are also well represented. 

The oxide sub-database is created using the FACT [12] database and contains the thermodynamic data for all solid oxides, sulphates and sulphides of the elements contained in the databases for Fe-based and Ni-based alloys. 

This tutorial, which is based on the Beilstein update BS0202PR (May, 2002) and on the retrieval program Cro.ssFire Commander V6,. shows. some typical advanced search examples in the Beilstein database. It is assumed that the user already knows some of the basic features of the retrieval program. Moreover, in this tutorial the CrossFire Structure Editor is used instead of the (SIS/Draw Structure Editor. The first example is a combined application of structure aiM fact retrieval, whereas the second example demonstrates reaction retrieval. 

Alternative technologies should also be considered, and the reasons for not using them should be justifiable. For example, database technology is not the right choice if a task requires reasoning that goes beyond retrieval of stored data based on well-defined criteria. At the same time, many problems that are stated in a symboHc way can be formulated mathematically, and in fact can be better solved numerically. For such problems, knowledge-based systems are not the appropriate answer. 

---