With the linguistic module EXTRAKT the, in all SISIS-SunRise query modules already implemented, quick and easy access to library stocks will be even more comfortable and higher in quality.
This new module analyses user input by linguistic methods, it supplements the query with terms from different kinds of dictionaries and passes those altered query strings on to concerning SISIS subsystems like SISIS-SunRise Catalogue, SISIS-SunRise webOPAC or SISIS-Elektra.
Linguistic search is activated by clicking a button within the query . Within the scope of analysis the query request is extended in following steps.
- Entered search terms are first of all examined and standardised on several different linguistic word variations (plural, singular, declensions, cases, etc.) using a standard dictionary (about 920.000 entries).
- In German often found composita are fragmented into its components using special dictionaries.
- Old and new orthographic forms are considered on demand.
- A synonyms dictionary can be optionally added.
- Some special thesauri (e.g. for towns, companies/organisations, abbreviations or acronyms) can also be integrated optionally.
- An optional multi-lingual search allows translating the query request into different languages, by which the query then takes place. Corresponding dictionaries are provided for German, English, French, Italian and Spanish. More planed languages are Dutch, Greek, Portuguese.
Example of a linguistically extended search:
It's looked for "Gartenlokal" AND "schattig".
Without linguistic support exactly those two terms will be queried in the SISIS indices.
With linguistic support, search criteria is extended to the following.
Breaking down composita adds the search terms „Garten“ and „Lokal“
By analysing variants of the term "schattig" supplemental terms "schattiger", "schattiges", "schattigem", "schattiger" are used as well.
On that a number of advantages arise for your portal visitors:
- The query is optimised without effort on the portal visitor.
- He is not concerned with alternative spellings, differing word forms, old or new orthography.
- The many options of the module make simple as well as special fields of application possible.
EXTRAKT is a product of our partner TEXTEC. Since the early 90ties TEXTEC uses linguistic methods for increasing quality of data retrieval.
EXTRAKT was initiated in some European Community projects.
