--- title: "Xiujun Li — author profile and papers" description: "Xiujun Li, Hebei University of Engineering: 4 papers on Elamur Research (2017–2020), h-index 37 — affiliations, topics, co-authors and who cites them in…" canonical: "https://research.elamur.ai/free/en/authors/xiujun-li/" --- [Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/authors/xiujun-li/) [home](https://md.research.elamur.ai/free/en/) / [authors](https://md.research.elamur.ai/free/en/authors/) / Xiujun Li # Xiujun Li Xiujun Li works on neural dialogue systems and, more recently, on pretrained models that connect language with images. Li's papers here, dated 2017 to 2020, ask how conversational agents can be trained end to end instead of as chains of separately built modules, how they can query structured databases, and how they can generate replies when little annotated data exists. The methods include reinforcement learning, simulated users and Transformer-based pretraining. Movie-ticket booking and multi-domain dialogue corpora serve as test settings. The most cited work is [Oscar](https://md.research.elamur.ai/free/en/papers/li-2020-oscar-object-semantics-aligned-pre/) (2020). It feeds a vision-language Transformer object tags predicted by a detector, alongside image regions and caption tokens. Because the tags are themselves words, they help the model match visual regions to text. [Towards End-to-End Reinforcement Learning of Dialogue Agents for Information Access](https://md.research.elamur.ai/free/en/papers/dhingra-2017-end-end-reinforcement-learning-dialogue/) (2017) replaces an exact database lookup with a probabilistic one. This keeps the parser's uncertainty and lets language understanding and policy be trained jointly. [End-to-End Task-Completion Neural Dialogue Systems](https://md.research.elamur.ai/free/en/papers/li-2017-end-end-task-completion-neural/) (2017) addresses errors that build up across pipeline stages. It trains a single neural system that consults a database to complete bookings. The 2020 papers move from reinforcement-learned dialogue agents toward large pretrained models. Besides Oscar, [Few-shot Natural Language Generation for Task-Oriented Dialog](https://md.research.elamur.ai/free/en/papers/peng-2020-few-shot-natural-language-generation/) introduces FewShotWOZ, a benchmark of seven domains with at most 50 training utterances each. The same paper presents SC-GPT, a GPT-2-based generator trained in stages, which turns dialog acts into fluent responses. The papers here fall mainly under [Dialogue systems](https://md.research.elamur.ai/free/en/topics/dialogue-systems/), with further work in [Vision & language](https://md.research.elamur.ai/free/en/topics/multimodal/). The most frequent co-author is [Jianfeng Gao](https://md.research.elamur.ai/free/en/authors/jianfeng-gao/). Other collaborators include [Lihong Li](https://md.research.elamur.ai/free/en/authors/lihong-li/) and [Yun-Nung Chen](https://md.research.elamur.ai/free/en/authors/yun-nung-chen/) on the dialogue papers, and Chunyuan Li. Bylines from 2019 onward include the University of Washington in Seattle. 4 papers on this site · 2017–2020 current affiliation [Hebei University of Engineering](https://ror.org/036h65h05), Handan, China; [Changchun University of Science and Technology](https://ror.org/007mntk44), Changchun, China; [Shandong University](https://md.research.elamur.ai/free/en/universities/shandong-university/), Jinan, China Xiujun Li has 4 papers on this site, in 2017–2020, most of them on [Dialogue systems](https://md.research.elamur.ai/free/en/topics/dialogue-systems/), [Language models & pretraining](https://md.research.elamur.ai/free/en/topics/language-models/), [Text generation](https://md.research.elamur.ai/free/en/topics/text-generation/); often with [Jianfeng Gao](https://md.research.elamur.ai/free/en/authors/jianfeng-gao/), [Chunyuan Li](https://md.research.elamur.ai/free/en/authors/chunyuan-li/), [Lihong Li](https://md.research.elamur.ai/free/en/authors/lihong-li/). [OpenAlex](https://openalex.org/A5021140826) · [ORCID](https://orcid.org/0000-0001-7771-2725) · [**h-index 37**](https://openalex.org/authors/A5021140826) · [278 works](https://openalex.org/works?filter=author.id:A5021140826) · [7 505 citations](https://openalex.org/works?filter=author.id:A5021140826&sort=cited_by_count:desc) OpenAlex, 2026-10-14 [4 works on this site](https://md.research.elamur.ai/free/en/authors/xiujun-li/#papers) · [609 citations in our library](https://md.research.elamur.ai/free/en/authors/xiujun-li/cited-by/) · 1 more work in our library 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=Xiujun+Li) · [Semantic Scholar](https://www.semanticscholar.org/search?q=Xiujun+Li) · [dblp](https://dblp.org/search?q=Xiujun+Li) · [ACL Anthology](https://aclanthology.org/search/?q=Xiujun+Li) ## Affiliations - [Shandong University](https://md.research.elamur.ai/free/en/universities/shandong-university/) · Jinan, China —current · 2001–2026 on 73 papers - [Changchun University of Science and Technology](https://ror.org/007mntk44) · Changchun, China —current · 2012–2026 on 36 papers - [Hebei University of Engineering](https://ror.org/036h65h05) · Handan, China —current · 2026 on 1 paper - [Chinese Academy of Sciences](https://md.research.elamur.ai/free/en/government/chinese-academy-of-sciences/) · Beijing, China —2003–2025 on 38 papers - [Northeast Institute of Geography and Agroecology](https://ror.org/01a9z1q73) · Changchun, China —2003–2025 on 38 papers - [University of Chinese Academy of Sciences](https://md.research.elamur.ai/free/en/universities/university-of-chinese-academy-of-sciences/) · Beijing, China —2018–2025 on 9 papers - [Shandong Agricultural University](https://ror.org/02ke8fw32) · Tai’an, China —2024–2025 on 2 papers - [State Grid Corporation of China (China)](https://ror.org/05twwhs70) · Beijing, China —2024–2025 on 2 papers - [Shanghai Normal University](https://ror.org/01cxqmw89) · Shanghai, China —2017–2024 on 11 papers - [University of Washington](https://md.research.elamur.ai/free/en/universities/university-of-washington/) · Seattle, United States —2019–2024 on 6 papers From the bylines of their papers (OpenAlex, 2026-10-14), ORCID and Wikidata; institutions on papers are matched by machine. 37 more institutions on single bylines are not listed. ## Topics [Dialogue systems 3](https://md.research.elamur.ai/free/en/topics/dialogue-systems/) · [Language models & pretraining 1](https://md.research.elamur.ai/free/en/topics/language-models/) · [Text generation 1](https://md.research.elamur.ai/free/en/topics/text-generation/) · [Vision & language 1](https://md.research.elamur.ai/free/en/topics/multimodal/) ## Works by year ### 2020s [**2020** 2 papers](https://md.research.elamur.ai/free/en/authors/xiujun-li/year/2020/) ### 2010s [**2017** 2 papers](https://md.research.elamur.ai/free/en/authors/xiujun-li/year/2017/) ## Papers The most cited first (OpenAlex and our library). 1. **[Oscar: Object-Semantics Aligned Pre-training for Vision-Language Tasks](https://md.research.elamur.ai/free/en/papers/li-2020-oscar-object-semantics-aligned-pre/)** *(2020 · 1 827 citations)* Vision-language pre-training (VLP) models usually feed image region features and text tokens together into a Transformer. 2. **[Towards End-to-End Reinforcement Learning of Dialogue Agents for Information Access](https://md.research.elamur.ai/free/en/papers/dhingra-2017-end-end-reinforcement-learning-dialogue/)** *(2017 · 347 citations)* The paper addresses how dialogue agents that help users find an entity in a Knowledge Base (KB-InfoBots) can be trained end-to-end. 3. **[Few-shot Natural Language Generation for Task-Oriented Dialog](https://md.research.elamur.ai/free/en/papers/peng-2020-few-shot-natural-language-generation/)** *(2020 · 253 citations)* Task-oriented dialog systems need an NLG component that turns a dialog act, meaning an intent plus slot-value pairs, into a fluent reply. 4. **[End-to-End Task-Completion Neural Dialogue Systems](https://md.research.elamur.ai/free/en/papers/li-2017-end-end-task-completion-neural/)** *(2017 · 141 citations)* Classic task-completion dialogue systems chain separately trained modules (LU, dialogue manager, NLG), so mistakes made early propagate and pile up downstream. ## Also in our library 1 more work of theirs is in our library without a page here yet; it will appear as we publish them. ## Co-authors here 12 of 17 co-authors with a page, the most papers together first. - [Jianfeng Gao](https://md.research.elamur.ai/free/en/authors/jianfeng-gao/) ×4 - [Chunyuan Li](https://md.research.elamur.ai/free/en/authors/chunyuan-li/) ×2 - [Lihong Li](https://md.research.elamur.ai/free/en/authors/lihong-li/) ×2 - [Yun-Nung Chen](https://md.research.elamur.ai/free/en/authors/yun-nung-chen/) ×2 - [Asli Celikyilmaz](https://md.research.elamur.ai/free/en/authors/asli-celikyilmaz/) - [Baolin Peng](https://md.research.elamur.ai/free/en/authors/baolin-peng/) - [Bhuwan Dhingra](https://md.research.elamur.ai/free/en/authors/bhuwan-dhingra/) - [Chenguang Zhu](https://md.research.elamur.ai/free/en/authors/chenguang-zhu/) - [Faisal Ahmed](https://md.research.elamur.ai/free/en/authors/faisal-ahmed/) - [Furu Wei](https://md.research.elamur.ai/free/en/authors/furu-wei/) - [Lei Zhang](https://md.research.elamur.ai/free/en/authors/lei-zhang/) - [Li Deng](https://md.research.elamur.ai/free/en/authors/li-deng/) ## Cited by Papers of our library cite their works 609 times (of the 140 884 papers whose bibliographies we read); [28 citing papers with a page here →](https://md.research.elamur.ai/free/en/authors/xiujun-li/cited-by/).