Refine
Document Type
- Conference Proceeding (6) (remove)
Keywords
- Automatische Spracherkennung (6) (remove)
Publicationstate
- Zweitveröffentlichung (6) (remove)
Reviewstate
- Peer-Review (5)
- (Verlags)-Lektorat (1)
Publisher
We present a fine-grained NER annotations scheme with 30 labels and apply it to German data. Building on the OntoNotes 5.0 NER inventory, our scheme is adapted for a corpus of transcripts of biographic interviews by adding categories for AGE and LAN(guage) and also adding label classes for various numeric and temporal expressions. Applying the scheme to the spoken data as well as a collection of teaser tweets from newspaper sites, we can confirm its generality for both domains, also achieving good inter-annotator agreement. We also show empirically how our inventory relates to the well-established 4-category NER inventory by re-annotating a subset of the GermEval 2014 NER coarse-grained dataset with our fine label inventory. Finally, we use a BERT-based system to establish some baselines for NER tagging on our two new datasets. Global results in in-domain testing are quite high on the two datasets, near what was achieved for the coarse inventory on the CoNLLL2003 data. Cross-domain testing produces much lower results due to the severe domain differences.
This paper presents experiments on sentence boundary detection in transcripts of spoken dialogues. Segmenting spoken language into sentence-like units is a challenging task, due to disfluencies, ungrammatical or fragmented structures and the lack of punctuation. In addition, one of the main bottlenecks for many NLP applications for spoken language is the small size of the training data, as the transcription and annotation of spoken language is by far more time-consuming and labour-intensive than processing written language. We therefore investigate the benefits of data expansion and transfer learning and test different ML architectures for this task. Our results show that data expansion is not straightforward and even data from the same domain does not always improve results. They also highlight the importance of modelling, i.e. of finding the best architecture and data representation for the task at hand. For the detection of boundaries in spoken language transcripts, we achieve a substantial improvement when framing the boundary detection problem as a sentence pair classification task, as compared to a sequence tagging approach.
The newest generation of speech technology caused a huge increase of audio-visual data nowadays being enhanced with orthographic transcripts such as in automatic subtitling in online platforms. Research data centers and archives contain a range of new and historical data, which are currently only partially transcribed and therefore only partially accessible for systematic querying. Automatic Speech Recognition (ASR) is one option of making that data accessible. This paper tests the usability of a state-of-the-art ASR-System on a historical (from the 1960s), but regionally balanced corpus of spoken German, and a relatively new corpus (from 2012) recorded in a narrow area. We observed a regional bias of the ASR-System with higher recognition scores for the north of Germany vs. lower scores for the south. A detailed analysis of the narrow region data revealed – despite relatively high ASR-confidence – some specific word errors due to a lack of regional adaptation. These findings need to be considered in decisions on further data processing and the curation of corpora, e.g. correcting transcripts or transcribing from scratch. Such geography-dependent analyses can also have the potential for ASR-development to make targeted data selection for training/adaptation and to increase the sensitivity towards varieties of pluricentric languages.
In diesem Beitrag wird das Redewiedergabe-Korpus (RW-Korpus) vorgestellt, ein historisches Korpus fiktionaler und nicht-fiktionaler Texte, das eine detaillierte manuelle Annotation mit Redewiedergabeformen enthält. Das Korpus entsteht im Rahmen eines laufenden DFG-Projekts und ist noch nicht endgültig abgeschlossen, jedoch ist für Frühjahr 2019 ein Beta-Release geplant, welches der Forschungsgemeinschaft zur Verfügung gestellt wird. Das endgültige Release soll im Frühjahr 2020 erfolgen. Das RW-Korpus stellt eine neuartige Ressource für die Redewiedergabe-Forschung dar, die in dieser Detailliertheit für das Deutsche bisher nicht verfügbar ist, und kann sowohl für quantitative linguistische und literaturwissenschaftliche Untersuchungen als auch als Trainingsmaterial für maschinelles Lernen dienen.
The administration of electronic publication in the Information Era congregates old and new problems, especially those related with Information Retrieval and Automatic Knowledge Extraction. This article presents an Information Retrieval System that uses Natural Language Processing and Ontology to index collection’s texts. We describe a system that constructs a domain specific ontology, starting from the syntactic and semantic analyses of the texts that compose the collection. First the texts are tokenized, then a robust syntactic analysis is made, subsequently the semantic analysis is accomplished in conformity with a metalanguage of knowledge representation, based on a basic ontology composed of 47 classes. The ontology, automatically extracted, generates richer domain specific knowledge. It propitiates, through its semantic net, the right conditions for the user to find with larger efficiency and agility the terms adapted for the consultation to the texts. A prototype of this system was built and used for the indexation of a collection of 221 electronic texts of Information Science written in Portuguese from Brazil. Instead of being based in statistical theories, we propose a robust Information Retrieval System that uses cognitive theories, allowing a larger efficiency in the answer to the users queries.