background image
call the Text Retrieval algorithm implemented in the database. Illustration 3 shows the
XML fragment of the Mapping Tree for the Emblem Book database depicting the map-
ping information associated to the attribute "Title."
I m p l e m e n t a t i o n o f t h e S y s t e m
This section shows how the system uses the Concept and Mapping Trees to build and execute
a query. Following this, the main modules of the system are described.
Query Interface Generator
We believe the user interface is a crucial aspect for the success of any computer system in gen-
eral, but particularly of any Web system.Therefore, our architecture is highly concerned with the
design of an intuitive, friendly, and easy-to-use interface.We have proposed three techniques to
design user interfaces that, being combined and systematically used, can improve the success of
the interface:
ˇ Use of Cognitive Metaphors:Web pages are built taking elements from the real world.This
technique is widely used, not only for the design of Web pages, but also for any computer
ˇ Use of Bounded Natural Language: It consists of offering the user a set of natural language
sentences with gaps.The user must choose the sentences of interest and fill the gaps in
order to express the query.
ˇ Navigational Approach:The main idea is to avoid the use of complex query languages to
obtain a very precise answer. Instead, we propose the use of an interface that guides the
user, in an intuitive manner, from a broad answer to a more precise one, progressively refin-
ing the query.
We also believe that having several levels of complexity for different user profiles is very use-
ful, so we offer unsophisticated interfaces to express simple queries for general users, and slight-
ly more complicated interfaces to allow experts to express more complex queries. The Query
Interface Generator offers the user a controlled natural language (Bounded Natural Language or
See N. R. Brisaboa, M. R Pen-
abad, A.S., Places, and F. J. Rodríguez,
"Tools for the Design of User-
Friendly Web Applications," Lecture
Notes in Computer Science 2115
(2001): 29-38, see also N.R. Bris-
aboa, M.R. Penabad, A.S. Places,
and F.J. Rodríguez,"A Document
Database Query Language," Lecture
Notes in Computer Science 2405
(2002): 183-198.
Illustration 2: A Concept Tree.
DC_Emblemsbook_180204 19.02.2004 11:26 Uhr Seite 102