Closed Mandalka closed 6 years ago
Automatic language detection by tika-python
Split Solr synonym config writing to different languages so only synonyms of the language are considered and language specific stemming works for them, too.
Since my projects have domain specific vocabularies with few false friends with the other languages, splitting synonyms config files to languages not necessary yet, since we can use only one and the same synonyms config for all different languages/stemming analysis.
Automatic language detection implemented.
Named Entities Recognition by Stanford NER now parametered with document language specific classifiers.
Open Semantic ETL now supports multilingual index for additional language specific analysis, i.e. for language specific grammar/stemming or synonyms
Search UI (Solr-PHP-UI) support the new multilingual search index structure: It searches in/with multiple/language specific fields/stemmers/grammars and provides language specific highlighting.
Added additional out of the box languages/grammars which are used by the multilingual research projects for stemming.