The Internet is a medium of two - way communication and continuously increases the user’s choices and the categories of demands to which he can respond. Practice so far has made quite commonplace the ability to have different options of appearance in a lot of web sites . The graphics , the frames , the colors, the fonts , the arrangement on the screen , the appearance or hiding of images and help tools, the enlargement or shrinking of the navigation map are the usual elements that are now organized – to lesser or greater extend – according to the preferences and the needs of the user. The rationale of increased initiatives on the part of the users presupposes a system of increased capabilities of the project. Capabilities, which however do not appear automatically, but are revealed gradually in different levels of penetration.

The fact that digital historical works are presented in a number of languages is yet another significant parameter, since we have increased accessibility with minimum cost. At the same time, it is more and more frequent to have web sites that take under consideration people of special needs (large fonts and voice creation systems).

By recording the habits and the interests of the user it creates a type of individual file of behavior, in order to facilitate as much as possible future visits. The recording of the searches, of their results and the suitability of the latter creates the opportunity (provided that it is a notional search) for improving the search procedure for any future user. The most common searches are based either on Free Text Retrieval (FTR) or on keywords that have been used in the texts under examination and usually allow the use of logical rules of the Boolean algebra. However, notional searches provide more accurate results and are more suitable for historical texts, since they take under consideration the different notional importance of the terms that are going to be searched (method of Vectorial Space Model - VSM) or the subjective structure of the texts (method of Latent Semantic Indexing or Latent Semantic Analysis - LSI ή LSA).