Profile avatar

Liu Jian

adherersun@163.com

Second-year Ph.D. Student, School of Artificial Intelligence, Beihang University
AI for Research, Recommender Systems, and World Models

Quick Facts

Research Interests: AI for Research, recommender systems, world models, large language models, and intelligent information processing.

Current Position: Second-year Ph.D. student at Beihang University, advised by Prof. Lei Sha.

About Me

I am Liu Jian, a second-year Ph.D. student in the School of Artificial Intelligence at Beihang University. I am advised by Prof. Lei Sha, and my current research interests lie in AI for Research, recommender systems, and world models.

Before starting my Ph.D. in September 2024, I worked as a Natural Language Processing Engineer at Baidu, where I focused on large language model data construction, document-level information extraction, low-resource SPO extraction, and healthcare knowledge engineering. This combination of academic and industrial experience continues to shape my research perspective on reliable, scalable, and practically useful AI systems.

I received my M.S. in Computer Science and Technology from the University of Electronic Science and Technology of China under the supervision of Prof. Zenglin Xu. Before that, I earned my B.S. in Computer Science and Technology from Fujian University of Technology, and I also studied as an exchange student at National Yunlin University of Science and Technology in Taiwan.

Download my English CV

Education

Beihang University

Ph.D. in Artificial Intelligence, School of Artificial Intelligence

Research Interests: AI for Research, recommender systems, and world models

Advisor: Prof. Lei Sha

Sep. 2024 - Present

University of Electronic Science and Technology of China

Academic M.S. in Computer Science and Technology

Research: low-resource NLP, domain adaptation, multi-task learning

Advisor: Prof. Zenglin Xu

Sep. 2018 - Jul. 2021

Fujian University of Technology

B.S. in Computer Science and Technology

GPA: 3.46 / 4.00

Exchange Study at National Yunlin University of Science and Technology, Taiwan

Sep. 2014 - Jun. 2018

Experience

Beihang University, School of Artificial Intelligence

Ph.D. Student

Advisor: Prof. Lei Sha

Sep. 2024 - Present

Conducting research on AI for Research, recommender systems, and world models, with a particular interest in building reliable and knowledge-intensive AI systems.

Baidu, Knowledge Graph Department

Natural Language Processing Engineer

Jul. 2021 - Apr. 2024

Worked on large language model data construction, document-level information extraction, low-resource SPO extraction, and healthcare knowledge engineering for real-world applications.

Research & Selected Projects

AI for Research. I am broadly interested in how AI systems can support scientific discovery, knowledge organization, evidence-grounded reasoning, and trustworthy research workflows.
Recommender Systems. I am interested in recommendation problems involving user modeling, representation learning, personalization, and scalable retrieval and ranking.
World Models. I am also interested in building models that can capture structure, dynamics, and long-horizon reasoning for intelligent decision-making and general-purpose AI systems.
ERNIE Bot Data Construction and Effectiveness Enhancement. At Baidu, I built and optimized vertical-domain datasets for large language models, including data construction, reasoning-chain resources, and evaluation workflows for practical deployment.
Document-level Information Extraction and Low-resource SPO Extraction. My previous work explored document-level extraction, prompt-based modeling, low-resource learning, and scalable NLP solutions for industrial knowledge engineering.

Publications

  1. Adversarial Training Based Multi-Source Unsupervised Domain Adaptation for Sentiment Analysis
    Yong Dai, Jian Liu, Xiancong Ren, and Zenglin Xu
    AAAI 2020 Poster
    Contribution: theory design for source-classifier weighting, co-training framework, benchmark experiments, implementation, and tuning.
  2. Unsupervised Sentiment Analysis by Transferring Multi-source Knowledge
    Yong Dai, Jian Liu, Jian Zhang, Hong-Bo Fu, and Zenglin Xu
    Cognitive Computation, 13 (2021): 1185-1197
    Contribution: overall model participation and benchmark experiment design, implementation, and tuning.
  3. Latent Dirichlet Allocation Based Generative Adversarial Networks
    Lili Pan, Shen Cheng, Jian Liu, Peijun Tang, Bowen Wang, Yazhou Ren, and Zenglin Xu
    Neural Networks
    Contribution: CIFAR-10 experiment tuning and experiment conclusion writing.

Awards & Honors

🏅
Baidu Excellent Project Award
Q1 2023, LLM reasoning data construction
🌟
Baidu Outstanding Newcomer
Q1 2022
🚀
Baidu Excellent Project Award
Q3 2021, industry graph construction migration efficiency
📘
Graduate Scholarships
UESTC, 2018 and 2019
🥇
National Encouragement Scholarship
and multiple university scholarships
💡
Patents
5 invention patents as first inventor, including 1 authorized patent

Skills

Research Areas: AI for Research, recommender systems, world models, large language models, information extraction, and knowledge-intensive AI.
Programming: Python, C / C++, algorithms, and data structures.
Frameworks: PyTorch, TensorFlow, and Keras.
Background: natural language processing, low-resource learning, domain adaptation, multi-task learning, and industrial AI system development.