--- title: "Shujie Liu — author profile and papers" description: "Shujie Liu, Microsoft: 5 papers on Elamur Research (2018–2023), h-index 45 — affiliations, topics, co-authors and who cites them in our library." canonical: "https://research.elamur.ai/free/en/authors/shujie-liu/" --- [Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/authors/shujie-liu/) [home](https://md.research.elamur.ai/free/en/) / [authors](https://md.research.elamur.ai/free/en/authors/) / Shujie Liu # Shujie Liu Shujie Liu works on natural language processing, with machine translation and large-scale pretraining at the centre. Liu's papers here, dated 2018 to 2023, repeatedly take methods developed for translation and language modelling and apply them to neighbouring problems: rewriting text in a different style, synthesising speech, representing source code and choosing replies in dialogue. Several of them also address evaluation, either by defining how a claim should be tested or by building a dataset that exposes a weakness in existing systems. The most cited work is [Achieving Human Parity on Automatic Chinese to English News Translation](https://md.research.elamur.ai/free/en/papers/hassan-2018-achieving-human-parity-automatic-chinese/) (2018). It turns "human parity" into a testable criterion. Bilingual judges rate machine output alongside professional translations of one shared test set, and parity holds when the gap between the two scores is not statistically significant. Judging against the source replaces reference-based metrics such as BLEU, and the paper describes the Chinese-to-English neural system that was assessed this way. In the same year, [Style Transfer as Unsupervised Machine Translation](https://md.research.elamur.ai/free/en/papers/zhang-2018-style-transfer-unsupervised-machine-translation/) treats two text styles, such as positive and negative sentiment, as if they were two languages. It uses phrase-based translation to produce rough pseudo-parallel sentence pairs, which neural models then refine, and an attribute classifier filters out pairs with the wrong style. The newer papers move into speech, code and dialogue. [Neural Codec Language Models are Zero-Shot Text to Speech Synthesizers](https://md.research.elamur.ai/free/en/papers/wang-2023-neural-codec-language-models-zero/) (2023) aims to reproduce voices never heard during training, without speaker adaptation or a separate speaker encoder. [GraphCodeBERT](https://md.research.elamur.ai/free/en/papers/guo-2021-graphcodebert-pre-training-code-representations/) (2021) pretrains a Transformer on code together with a graph that traces how values pass between variables. [MuTual](https://md.research.elamur.ai/free/en/papers/cui-2020-mutual-dataset-multi-turn-dialogue/) (2020) is a hand-annotated set of 8,860 multi-turn dialogues that tests whether next-utterance prediction involves real reasoning rather than surface matching. Liu is affiliated with [Microsoft](https://md.research.elamur.ai/free/en/companies/microsoft/) and published from [Microsoft Research Asia](https://md.research.elamur.ai/free/en/companies/microsoft-research-asia/) in Beijing between 2013 and 2025. Co-authors on the papers here include [Ming Zhou](https://md.research.elamur.ai/free/en/authors/ming-zhou/), [Mu Li](https://md.research.elamur.ai/free/en/authors/mu-li/), Long Zhou, Shuo Ren, Yu Wu and [Hany Hassan](https://md.research.elamur.ai/free/en/authors/hany-hassan/). The work falls mainly under [machine translation](https://md.research.elamur.ai/free/en/topics/machine-translation/) and language models and pretraining. Shujie Liu has 5 papers in our library, in 2018–2023, most of them on [Machine translation](https://md.research.elamur.ai/free/en/topics/machine-translation/), [Language models & pretraining](https://md.research.elamur.ai/free/en/topics/language-models/), [Dialogue systems](https://md.research.elamur.ai/free/en/topics/dialogue-systems/); often with [Long Zhou](https://md.research.elamur.ai/free/en/authors/long-zhou/), [Ming Zhou](https://md.research.elamur.ai/free/en/authors/ming-zhou/), [Mu Li](https://md.research.elamur.ai/free/en/authors/mu-li/). 5 papers in our library · 2018–2023 [OpenAlex](https://openalex.org/A5101635405) · [ORCID](https://orcid.org/0009-0008-2599-6752) · [**h-index 45**](https://openalex.org/authors/A5101635405) · [288 works](https://openalex.org/works?filter=author.id:A5101635405) · [10,428 citations](https://openalex.org/works?filter=author.id:A5101635405&sort=cited_by_count:desc) OpenAlex, 2026-10-14 [📄 5 works in our library](https://md.research.elamur.ai/free/en/authors/shujie-liu/#papers) · [🔗 865 citations in our library](https://md.research.elamur.ai/free/en/authors/shujie-liu/cited-by/) · [5 more in our library without a page yet](#also) · [⭐ Paper Star rank #1289](https://md.research.elamur.ai/free/en/paper-stars/page/13/#star-1289) The profile is the library's match of the name; two people with one name are not always told apart. Elsewhere: [Google Scholar](https://scholar.google.com/citations?view_op=search_authors&mauthors=Shujie+Liu) · [Semantic Scholar](https://www.semanticscholar.org/search?q=Shujie+Liu) · [dblp](https://dblp.org/search?q=Shujie+Liu) · [ACL Anthology](https://aclanthology.org/search/?q=Shujie+Liu) ## Affiliations [🏢**Microsoft** Redmond · United States 2013–2026 · 66 papers*current*](https://md.research.elamur.ai/free/en/companies/microsoft/) · [🏢**Powerchina Huadong Engineering Corporation** Hangzhou · China 2022–2026 · 2 papers*current*](https://ror.org/03mcefb58) · [🎓**Shijiazhuang University** Shijiazhuang · China 2023–2026 · 2 papers*current*](https://ror.org/028rmam09) · [🎓**Changchun University** Changchun · China 2026 · 1 paper*current*](https://ror.org/02an57k10) · [🏢**China Huaneng Group Co., Ltd.** Beijing · China 2026 · 1 paper*current*](https://ror.org/059wz5438) · [🏢**Microsoft Research Asia** Beijing · China 2013–2025 · 54 papers](https://md.research.elamur.ai/free/en/companies/microsoft-research-asia/) · [🎓**Dalian University of Technology** Dalian · China 2013–2025 · 29 papers](https://md.research.elamur.ai/free/en/universities/dalian-university-of-technology/) · [🏢**China National Offshore Oil Corporation** Beijing · China 2009–2025 · 15 papers](https://ror.org/054dq0621) · [🎓**Qufu Normal University** Qufu · China 2013–2023 · 15 papers](https://ror.org/03ceheh96) · [🎓**Harbin Institute of Technology** Harbin · China 2006–2023 · 13 papers](https://md.research.elamur.ai/free/en/universities/harbin-institute-of-technology/) From the bylines of their papers (OpenAlex, 2026-10-14), ORCID and Wikidata; institutions on papers are matched by machine. 26 more institutions on single bylines are not listed. ## Topics [Machine translation (2)](https://md.research.elamur.ai/free/en/topics/machine-translation/) · [Language models & pretraining (2)](https://md.research.elamur.ai/free/en/topics/language-models/) · [Dialogue systems (1)](https://md.research.elamur.ai/free/en/topics/dialogue-systems/) · [Speech & spoken language (1)](https://md.research.elamur.ai/free/en/topics/speech/) · [Evaluation & benchmarks (1)](https://md.research.elamur.ai/free/en/topics/evaluation-benchmarks/) · [Reasoning (1)](https://md.research.elamur.ai/free/en/topics/reasoning/) · [Code & programming languages (1)](https://md.research.elamur.ai/free/en/topics/code-and-programming/) ## Works by year ### 2020s [**2023** 1 paper](https://md.research.elamur.ai/free/en/papers/wang-2023-neural-codec-language-models-zero/) · [**2021** 1 paper](https://md.research.elamur.ai/free/en/papers/guo-2021-graphcodebert-pre-training-code-representations/) · [**2020** 1 paper](https://md.research.elamur.ai/free/en/papers/cui-2020-mutual-dataset-multi-turn-dialogue/) ### 2010s [**2018** 2 papers](https://md.research.elamur.ai/free/en/authors/shujie-liu/year/2018/) ## Papers The most cited first (OpenAlex and our library). 1. **[Achieving Human Parity on Automatic Chinese to English News Translation](https://md.research.elamur.ai/free/en/papers/hassan-2018-achieving-human-parity-automatic-chinese/)** *(2018 · 794 citations)* The paper asks whether machine translation can match professional human translators, and first makes that question testable. 2. **[Neural Codec Language Models are Zero-Shot Text to Speech Synthesizers](https://md.research.elamur.ai/free/en/papers/wang-2023-neural-codec-language-models-zero/)** *(2023 · 333 citations)* Zero-shot text-to-speech means reproducing a voice the system never saw during training. Earlier approaches relied on speaker adaptation or dedicated speaker encoders. Models trained on small, studio-clean corpora tend to generalize poorly to new speakers. 3. **[GraphCodeBERT: Pre-training Code Representations with Data Flow](https://md.research.elamur.ai/free/en/papers/guo-2021-graphcodebert-pre-training-code-representations/)** *(2021 · 328 citations)* Most pre-trained models for source code read a program as a plain token sequence. This ignores structural information that shows what the code actually does. GraphCodeBERT adds data flow to the input: a graph in which variables are nodes and directed edges show which variables each value is derived from. The authors… 4. **[MuTual: A Dataset for Multi-Turn Dialogue Reasoning](https://md.research.elamur.ai/free/en/papers/cui-2020-mutual-dataset-multi-turn-dialogue/)** *(2020 · 189 citations)* Open-domain chatbots often produce replies that are fluent and on topic but logically wrong. Existing response-selection benchmarks rarely expose this weakness, because surface matching solves most of their examples. The authors built MuTual, a manually annotated dataset of 8,860 multi-turn dialogues that tests… 5. **[Style Transfer as Unsupervised Machine Translation](https://md.research.elamur.ai/free/en/papers/zhang-2018-style-transfer-unsupervised-machine-translation/)** *(2018 · 164 citations)* Text style transfer means rewriting a sentence so that it carries a different attribute, such as positive instead of negative sentiment, while what it says stays the same. ## Also in our library 5 more works of theirs are in our library without a page here yet; they will appear as we publish them. ## Co-authors here 12 of 21 co-authors with a page, the most papers together first. - [Long Zhou](https://md.research.elamur.ai/free/en/authors/long-zhou/) ×2 - [Ming Zhou](https://md.research.elamur.ai/free/en/authors/ming-zhou/) ×2 - [Mu Li](https://md.research.elamur.ai/free/en/authors/mu-li/) ×2 - [Shuo Ren](https://md.research.elamur.ai/free/en/authors/shuo-ren/) ×2 - [Yu Wu](https://md.research.elamur.ai/free/en/authors/yu-wu/) ×2 - [Christian Federmann](https://md.research.elamur.ai/free/en/authors/christian-federmann/) - [Daya Guo](https://md.research.elamur.ai/free/en/authors/daya-guo/) - [Duyu Tang](https://md.research.elamur.ai/free/en/authors/duyu-tang/) - [Enhong Chen](https://md.research.elamur.ai/free/en/authors/enhong-chen/) - [Hany Hassan](https://md.research.elamur.ai/free/en/authors/hany-hassan/) - [Jonathan H. Clark](https://md.research.elamur.ai/free/en/authors/jonathan-h-clark/) - [Leyang Cui](https://md.research.elamur.ai/free/en/authors/leyang-cui/) ## Cited by Papers of our library cite their works 865 times (of the 140,884 papers whose bibliographies we read); [84 citing papers with a page here →](https://md.research.elamur.ai/free/en/authors/shujie-liu/cited-by/).