More and more companies provide their information in electronic form today,
and it can be found in large commercial databases or on the web. The probl
em with information has changed, however, from too little to too much infor
mation. The annual reports for example contain information both in numerica
l and symbolic form. So far only the numerical information has been analyze
d with help of computers. However, it is clear that the verbal information
is also of great value. Our paper aims at investigating the potential of co
mputer aided analysis of text. Our approach is based on smart encoding, sel
f-organizing maps, and on document histograms.