Hier finden Sie wissenschaftliche Publikationen aus den Fraunhofer-Instituten.

Hybrid ensemble predictor as quality metric for German text summarization: Fraunhofer IAIS at GermEval 2020 task 3

: Biesner, D.; Brito, E.; Hillebrand, L.P.; Sifa, R.

Volltext ()

Ebling, S.:
5th Swiss Text Analytics Conference and 16th Conference on Natural Language Processing, SWISSTEXT and KONVENS 2020. Proceedings. Online resource : Zurich, Switzerland, June 23-25, 2020 (held online due to COVID19 pandemic)
Aachen: CEUR-WS, 2020 (CEUR Workshop Proceedings 2624)
ISSN: 1613-0073
5 S.
Swiss Text Analytics Conference (SwissText) <5, 2020, Online>
Conference on Natural Language Processing (KONVENS) <16, 2020, Online>
Konferenzbeitrag, Elektronische Publikation
Fraunhofer IAIS ()

We propose an alternative quality metric to evaluate automatically generated texts based on an ensemble of different scores, combining simple rule-based metrics with more complex models of very different nature, including ROUGE, tf-idf, neural sentence embeddings, and a matrix factorization method. Our approach achieved one of the top scores on the second German Text Summarization Challenge.