Junxian Zhou (周俊贤)

*Algorithm Engineer · Applied NLP & LLM Systems*

I build language AI systems that bridge research and production, with a focus on natural language processing, information extraction, sentiment analysis, and the productionization of LLM-driven applications.

Seven years in applied algorithm engineering — three of them on LLM systems.

Research Interests / 研究方向

• LLM Systems: Application engineering, evaluation, productionization
• NLP: Aspect-based Sentiment Analysis, Information Extraction

Selected Projects / 开源项目

• python-node-deepresearch !Stars — A DeepResearch-style LLM agent for multi-step web research.
• NLP-Series-sentence-embeddings !Stars — Chinese sentence embeddings: BERT-avg, BERT-whitening, SBERT, SimCSE.
• NLP-Series-text-cls !Stars — Text classification baselines: BERT, semi-supervised UDA, adversarial training, data augmentation.
• NLP-Series-relation-extraction !Stars — Relation extraction: sequence tagging, cascaded pointer networks, multi-head selection, deep biaffine attention.
• DuEE !Stars — PyTorch baseline for the Baidu 2021 LIC event extraction track.
• NLP-Data-Augmentation !Stars — Text augmentation: synonym replacement (via word2vec) and back-translation.

Selected Publications / 学术发表

• J. Zhou, H. Yang, Y. He, H. Mou, J. Yang. *A Unified One-Step Solution for Aspect Sentiment Quad Prediction.* In *Findings of ACL 2023*, pp. 12249–12265. [[paper]](https://doi.org/10.18653/v1/2023.findings-acl.777)
• J. Zhou, H. Yang, Y. Junpeng, Y. He, H. Mou. *An Empirical Study of Benchmarking Chinese Aspect Sentiment Quad Prediction.* arXiv:2311.01713, 2023. [[paper]](https://doi.org/10.48550/arXiv.2311.01713)
• Z. R. Lu, Z. Yin, J. Zhou, J. Liu, L. Wang. *A simple and effective Measurement-Changes-Correction strategy for damage identification with aleatoric and epistemic model errors.* *Structural Health Monitoring*, 20(3): 1196–1220, 2021. [[paper]](https://doi.org/10.1177/1475921720948207)

Earlier publications

• Z. R. Lu, J. Zhou, L. Wang, J. Liu. *Damage identification from static tests by eigenparameter decomposition and sparse regularization.* *Structural Health Monitoring*, 19(5): 1351–1374, 2020. [[paper]](https://doi.org/10.1177/1475921719880980)
• L. Wang, J. Zhou, Z. R. Lu. *A fast friction-model-inspired sparse regularization approach for damage identification with modal data.* *Computers & Structures*, 227: 106142, 2020. [[paper]](https://doi.org/10.1016/j.compstruc.2019.106142)
• Z. R. Lu, J. Zhou, L. Wang. *On choice and effect of weight matrix for response sensitivity-based damage identification with measurement and model errors.* *Mechanical Systems and Signal Processing*, 114: 1–24, 2019. [[paper]](https://doi.org/10.1016/j.ymssp.2018.05.007)

Awards / 获奖

• 阿里巴巴 Create@AI 黑客松, 亚军, 2024.
• 天池"万创杯"中医药大数据挑战赛, 亚军, 2020.

Patent / 专利

• 周俊贤, 朱汝维. *一种利用命名实体识别提取营业执照结构化信息的方法.* CN112668335A, 2021.

Connect / 联系

• WeChat: ZHOU-JXX
• X (Twitter): @zhoujx4fox35923
• Google Scholar: Profile
• 知乎 · 周星星
• 小红书 · 周星星的AI煎饼摊

GitHub Stats

Collapse