Information Modelling and Retrieval in Hypermedia Systems

In the first generation of hypermedia systems the main focus was on the flexibility of information organisation and on the ease of navigational access. In particular, this way of interaction was emphasised in contrast with traditional information retrieval systems based on formal query languages. Unfortunately, the limitations of the browsing paradigm were soon realised when approaching real and large hypermedia applications and thus the well-known problems of disorientation and cognitive overhead have been frequently reported and discussed in the literature [18]. In order to alleviate such problems, search capabilities for hypertexts have been added in a more or less integrated way. The effect of this is that content-based requests can be submitted and directly evaluated yielding an immediate answer to a query, whereas a browsing session can take a long time before converging to the goal or it may not converge at all. The problem still remains that such queries do not take into account the structure of information and do not retrieve linked collections of information based on the specification of their structure as well as of their content.