Ming Gong

Ming Gong

Auto-generated profile, based on publicly available information

Ming Gong

Microsoft Corporation, Beijing, China

Research Focus & Metrics

Academic Career Length

8Years (2019-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
8 October 2023
Verified
Book Chapter
Closed access
PDF file
NWNing Wu
Ning Wu
MGMing Gong
Ming Gong
LSLinjun Shou
Linjun Shou
SLShining Liang
Shining Liang
DJDaxin Jiang
Daxin Jiang

9 May 2023 in ACM Transactions on Asian and Low-Resource Language Information Processing
Verified
Journal Article
Open access
JLJunwei Liao
Junwei Liao
SESefik Emre Eskimez
Sefik Emre Eskimez
LLLiyang Lu
Liyang Lu
YSYu Shi
Yu Shi
MGMing Gong
Ming Gong
MZMichael Zeng
Michael Zeng
Modern Automatic Speech Recognition (ASR) systems can achieve high performance in terms of recognition accuracy. However, a perfectly accurate transcript still can be challenging to read due to grammatical errors, disfluency, and other noises common in spoken communication. These readable issues introduced by speakers and ASR systems will impair the performance of downstream tasks and the understanding of human readers. In this work, we present a task called ASR post-processing for readability (APR) and formulate it as a sequence-to-sequence text generation problem. The APR task aims to transform the noisy ASR output into a readable text for humans and downstream tasks while maintaining the semantic meaning of speakers. We further study the APR task from the benchmark dataset, evaluation metrics, and baseline models: First, to address the lack of task-specific data, we propose a method to construct a dataset for the APR task by using the data collected for grammatical error correction. Second, we utilize metrics adapted or borrowed from similar tasks to evaluate model performance on the APR task. Lastly, we use several typical or adapted pre-trained models as the baseline models for the APR task. Furthermore, we fine-tune the baseline models on the constructed dataset and compare their performance with a traditional pipeline method in terms of proposed evaluation metrics. Experimental results show that all the fine-tuned baseline models perform better than the traditional pipeline method, and our adapted RoBERTa model outperforms the pipeline method by 4.95 and 6.63 BLEU points on two test sets, respectively. The human evaluation and case study further reveal the ability of the proposed model to improve the readability of ASR transcripts.


5 August 2020
Verified
Conference Paper
Closed access
YZYu Zeng
Yu Zeng
YQYong Qin
Yong Qin
DLDian Liu
Dian Liu
YFYong Fu
Yong Fu
MGMing Gong
Ming Gong
XZXinning Zhang
Xinning Zhang
The safe operation of railway train is very important to the safety of the railway system. Therefore, it is necessary to master the failure rules and causes of the railway train devices. However, the current common fault analysis methods have some shortcomings, such as complicated compilation process and insufficient expression. Therefore, we choose to use the knowledge graph technology. Firstly, the common faults of railway train devices are obtained through the statistics of real fault data; secondly, the causes of faults are analyzed and the fault tree is established; thirdly, the ontology of railway train system is modeled according to the idea of fault tree construction; finally, the logical relationship between ontology is established according to the ontology model, and the railway train device fault causality model based on knowledge graph is established.

26 November 2023
Verified
Conference Paper
Open access
SZShengyao Zhuang
Shengyao Zhuang
LSLinjun Shou
Linjun Shou
JPJian Pei
Jian Pei
MGMing Gong
Ming Gong
HRHouxing Ren
Houxing Ren
DJDaxin Jiang
Daxin Jiang
Current dense retrievers (DRs) are limited in their ability to effectively process misspelled queries, which constitute a significant portion of query traffic in commercial search engines. The main issue is that the pre-trained language model-based encoders used by DRs are typically trained and fine-tuned using clean, well-curated text data. Misspelled queries are typically not found in the data used for training these models, and thus misspelled queries observed at inference time are out-of-distribution compared to the data used for training and fine-tuning. Previous efforts to address this issue have focused on fine-tuning strategies, but their effectiveness on misspelled queries remains lower than that of pipelines that employ separate state-of-the-art spell-checking components. To address this challenge, we propose ToRoDer (TypOs-aware bottlenecked pre-training for RObust DEnse Retrieval), a novel pre-training strategy for DRs that increases their robustness to misspelled queries while preserving their effectiveness in downstream retrieval tasks. ToRoDer utilizes an encoder-decoder architecture where the encoder takes misspelled text with masked tokens as input and outputs bottlenecked information to the decoder. The decoder then takes as input the bottlenecked embeddings, along with token embeddings of the original text with the misspelled tokens masked out. The pre-training task is to recover the masked tokens for both the encoder and decoder. Our extensive experimental results and detailed ablation studies show that DRs pre-trained with ToRoDer exhibit significantly higher effectiveness on misspelled queries, sensibly closing the gap with pipelines that use a separate, complex spell-checker component, while retaining their effectiveness on correctly spelled queries.

1 November 2024
Verified
Book Chapter
Closed access
PDF file
ZXZilin Xiao
Zilin Xiao
MGMing Gong
Ming Gong
PCPaola Cascante-Bonilla
Paola Cascante-Bonilla
XZXingyao Zhang
Xingyao Zhang
JWJie Wu
Jie Wu
VOVicente Ordonez
Vicente Ordonez