You can not select more than 25 topics
Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
33 lines
1.6 KiB
33 lines
1.6 KiB
13 years ago
|
A vocabulary is used to produce search navigation entities.
|
||
|
|
A search navigation is what you see at the right column
|
||
|
|
at the side of a search results where it is possible to reduce the
|
||
|
|
set of result entries with given restrictions.
|
||
|
|
|
||
|
|
A vocabulary is a restriction where the search results are restricted to
|
||
|
|
entries which have a specific tag in the subject metadata that corresponds
|
||
|
|
to the vocabulary restriction. The restriction is expressed with a set of
|
||
|
|
synonyms for the tag in a property-like file. Such files are activated if
|
||
|
|
they are present in the folder DATA/DICTIONARIES/autotagging/ at start-up time
|
||
|
|
and the vocabulary files must be named with a '.vocabulary' extension.
|
||
|
|
|
||
|
|
Vocabulary files are similar to property-files with these rules rules:
|
||
|
|
- the key represents the vocabulary term (this is what you see in the navigation)
|
||
|
|
- a value is a list of synonyms for the vocabulary term
|
||
|
|
- a term is always self-referencing (the term is also a synonym for the term)
|
||
|
|
- a value may be omitted (a self-referencing-only vocabulary)
|
||
|
|
|
||
|
|
The format of a vocabulary file is:
|
||
|
|
each line has the format
|
||
|
|
<print-name>[=<synonym>{','<synonym>}*]
|
||
|
|
or the line starts with a '#' for comment lines
|
||
|
|
|
||
|
|
The subdirectories of this directory contains example-vocabularies for
|
||
|
|
specific languages. Vocabularies work best if the vocabulary is expressed in
|
||
|
|
the same language as the documents are that are indexed.
|
||
|
|
|
||
|
|
A vocabulary can be activated by doing:
|
||
|
|
- copy the vocabulary from the <lang>/ subdirectory to DATA/DICTIONARIES/autotagging/
|
||
|
|
- restart
|
||
|
|
- do an indexing of the web-pages. Vocabularies cannot be applied to already indexed
|
||
|
|
web pages because tags are only generated during the parsing process
|