Incorporating Pinyin into Pipeline Named Entity Recognition from Chinese Speech
Authors:
- 31 October 2023
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 947-953
Abstract
Named Entity Recognition (NER) from speech is usually implemented through a two-step pipeline that consists of (1) processing the audio using an Automatic Speech Recognition (ASR) system and (2) applying an NER tagger to the ASR output. In this paper, we incorporate pinyin 1 — spelled sounds of Chinese characters — into the pipeline NER from Chinese speech, aiming to improve the NER performance through two steps. First, we take the pretrained model ChineseBERT to embed pinyin, as well as glyph and char, as the input of NER tagger. Second, we introduce homophone noises into training data for NER tagger, as homophone errors most likely exist in ASR output for Chinese speech. Using the two-step pipeline method with pinyin incorporated into the NER tagger, the F1 score is improved by nearly 1% absolute points in the experiment on the AISHELL-NER dataset, which is a significant improvement in the field of NER. And the F1 score outperforms the current state-of-the-art (SOTA) result on the AISHELL-NER dataset by 0.4% absolute points, despite the slightly worse Character Error Rate (CER) of our ASR.Keywords
This publication has 15 references indexed in Scilit:
- WeNet: Production Oriented Streaming and Non-Streaming End-to-End Speech Recognition Toolkit2021
- SpanNER: Named Entity Re-/Recognition as Span Prediction2021
- ChineseBERT: Chinese Pretraining Enhanced by Glyph and Pinyin Information2021
- End-to-End Named Entity Recognition from English Speech2020
- ContextNet: Improving Convolutional Neural Networks for Automatic Speech Recognition with Global Context2020
- Conformer: Convolution-augmented Transformer for Speech Recognition2020
- Transformer Transducer: A Streamable Speech Recognition Model with Transformer Encoders and RNN-T Loss2020
- The Speechtransformer for Large-scale Mandarin Chinese Speech Recognition2019
- End-To-End Named Entity And Semantic Concept Extraction From Speech2018
- AISHELL-1: An open-source Mandarin speech corpus and a speech recognition baseline2017