| Title | Context based number normalization using skip-chain conditional random fields / |
|---|---|
| Authors | Balčiūnas, Linas |
| Full Text |
|
| Is Part of | CEUR workshop proceedings: IVUS 2019 international conference on information technologies: proceedings of the international conference on information technologies, Kaunas, Lithuania, April 25, 2019 / edited by: Robertas Damaśevićius, Tomas Krilavićius, Audrius Lopata, Dawid Połap, Marcin Woźniak.. Aachen : CEUR-WS. 2019, vol. 2470, p. 17-21.. ISSN 1613-0073 |
| Keywords [eng] | conditional random field ; natural language processing ; number normalization ; text normalization |
| Abstract [eng] | Verbalizing numeric text tokens is a required task for various speech-related applications, including automatic speech recognition and text-to-speech synthesis. In morphologically rich languages, such conversion involves predicting implicit morphological properties of a corresponding numeral. In this paper, we propose first-order skip-chain Conditional Random Field (CRF) models and various prepossessing techniques to leverage different contextual information. We show that our best skip-chain CRF models achieve over 80% accuracy on the set of 2000 Lithuanian sentences. |
| Published | Aachen : CEUR-WS |
| Type | Conference paper |
| Language | English |
| Publication date | 2019 |