Relevant links for message retrieval are not confined to synonymy.
The conceptualization of "reading" as an active process of sign-making, and not just information retrieval, supports both creative and oppositional meaning making.
The ontology is organized in simple taxonomies, and concepts from the taxonomy are then identified within the documents to enable the retrieval process.
Furthermore, while document processing concentrates on paragraph retrieval in the former instance, reference resolution throughout the entire text is paramount in the latter application.
Figure 21 shows an example of shape-based reminding or query by sketch retrieval of several visual and nonvisual databases.
Based on this analysis, we propose a method to select promising index terms from among various types of index terms in the second-phase retrieval.
Attempts to exploit word-sense identification for text retrieval have produced many negative findings and relatively few positive ones.
In this system, after the retrieval of the relevant cases, they are ranked heuristically by ease of adaptation.
