paper-with-me

홈 › Papers

DISCO: A Large Scale Human Annotated Corpus for Disfluency Correction in Indo-European Languages

2023-10-25 · Vineet Bhat, Preethi Jyothi, Pushpak Bhattacharyya

Disfluency correction (DC) is the process of removing disfluent elements like fillers, repetitions and corrections from spoken utterances to create readable and interpretable text. DC is a vital post-processing step applied to Automatic Speech Recognition (ASR) outputs, before subsequent processing by downstream language understanding tasks. Existing DC research has primarily focused on English due to the unavailability of large-scale open-source datasets. Towards the goal of multilingual disfluency correction, we present a high-quality human-annotated DC corpus covering four important Indo-European languages: English, Hindi, German and French. We provide extensive analysis of results of state-of-the-art DC models across all four languages obtaining F1 scores of 97.55 (English), 94.29 (Hindi), 95.89 (German) and 92.97 (French). To demonstrate the benefits of DC on downstream tasks, we show that DC leads to 5.65 points increase in BLEU scores on average when used in conjunction with a state-of-the-art Machine Translation (MT) system. We release code to run our experiments along with our annotated dataset here.

📄 PDF Abstract BibTeX arXiv:2310.16749

Code (1)

vineet2104/disco 공식 구현

Tasks

Automatic Speech RecognitionAutomatic Speech Recognition (ASR)Machine Translationspeech-recognitionSpeech Recognition

Similar Papers 제목 키워드 기반

MEGA RST Discourse Treebanks with Structure and Nuclearity from Scalable Distant Sentiment Supervision

2020-11-05 · EMNLP 2020 11 · Patrick Huber, Giuseppe Carenini

The lack of large and diverse discourse treebanks hinders the application of data-driven approaches, such as deep-learning, to RST-style discourse parsing. In this work, we present a novel scalable methodology to automat…

Discourse Parsing

Relation Discovery with Out-of-Relation Knowledge Base as Supervision

2019-04-19 · NAACL 2019 6 · Yan Liang, Xin Liu, Jianwen Zhang, Yangqiu Song

Unsupervised relation discovery aims to discover new relations from a given text corpus without annotated data. However, it does not consider existing human annotated knowledge bases even when they are relevant to the re…

Relation

RAAMove: A Corpus for Analyzing Moves in Research Article Abstracts

2024-03-23 · Hongzheng Li, Ruojin Wang, Ge Shi, Xing Lv 외

Move structures have been studied in English for Specific Purposes (ESP) and English for Academic Purposes (EAP) for decades. However, there are few move annotation corpora for Research Article (RA) abstracts. In this pa…

A Bilingual Parallel Corpus with Discourse Annotations

2022-10-26 · Yuchen Eleanor Jiang, Tianyu Liu, Shuming Ma, Dongdong Zhang 외

Machine translation (MT) has almost achieved human parity at sentence-level translation. In response, the MT community has, in part, shifted its focus to document-level translation. However, the development of document-l…

Document Level Machine TranslationMachine TranslationSentenceTranslation

A Novel Corpus of Discourse Structure in Humans and Computers

2021-11-10 · Babak Hemmatian, Sheridan Feucht, Rachel Avram, Alexander Wey 외

We present a novel corpus of 445 human- and computer-generated documents, comprising about 27,000 clauses, annotated for semantic clause types and coherence relations that allow for nuanced comparison of artificial and n…

Text Generation