-
Notifications
You must be signed in to change notification settings - Fork 0
Home
runnwerth edited this page Feb 25, 2021
·
38 revisions
In diesem Repositorium probieren wir annif mit den Daten der TIB aus.
Im Folgenden bereiten wir Daten für annif auf, verwenden annif zur automatischen Indexierung und diskutieren die Ergebnisse.
- Übersetzen der Vokabulardaten in die annif-Syntax
- Daten-Dump festlegen
- Daten Dump herunterladen
- Übersetzen des Dumps in die annif-Syntax
- Anlegen eines Goldstandards
- Extrahieren einer zufälligen Testmenge
- Anlegen eines annif-Projektes
-
Ergebnisse
- Versuch 1: TIBKAT-Metadaten mit Basisklassifikation und Dokumentsprache deutsch, Omikuji
- Versuch 2: TIBKAT-Metadaten mit Abstracts, GND-Schlagwörtern und Dokumentsprache deutsch, Omikuji
- Versuch 3: TIBKAT-Metadaten mit Abstracts, GND-Schlagwörtern und Dokumentsprache englisch, Omikuji
- Versuch 4: FIDmove-Index (TIBKAT-Anteil), GND-Schlagwörter, deutsch, Omikuji
- Versuch 5: FIDmove-Index (TIBKAT-Anteil), GND-Schlagwörter, deutsch, Maui
- Versuch 6: TIBKAT, Lokalsystematik Mathematik, deutsch, Omikuji
- Versuch 7: TIBKAT, LinSearch, deutsch
- Versuch 8: TIBKAT, Lokalsystematik Rechtswissenschaften, deutsch
- Weitere Ideen
- Veröffentlichungen