Dian Chen
高级算法工程师 at 阿里巴巴（北京）软件服务有限公司
Haidian District, Beijing, China
Summary
👤
Senior🎓
Top SchoolDian Chen is a senior algorithm engineer with 11 years of experience specializing in recommender systems and cognitive/interactive recommendation at Alibaba in Beijing. He led the design and delivery of RecGPT-based cognitive recall and recommendation pipelines, launched multiple productized models that materially improved discovery metrics, and spun up new teams to operationalize these capabilities. His background blends deep academic research—a PhD in data mining and machine learning from UCAS—with hands-on engineering across NLP tasks such as text error correction, entity extraction, and Text2SQL, evidenced by contributions to the widely used pycorrector toolkit. Dian has repeatedly turned research prototypes into production services for finance and e-commerce, authored ATA technical docs, and won internal hackathon recognition for interactive recommendation designs. Equally comfortable with model pretraining and inference engineering, he pays close attention to practical issues like UNK token alignment and OOV handling that often make or break production performance. Based in Haidian, he pairs rigorous scientific training with a consistent track record of shipping scalable ML systems that improve user discovery.
11 years of coding experience
工学学士, 物联网专业, GPA 3.21+ / 3.60+, 工学学士, 物联网专业, GPA 3.21+ / 3.60+ at 重庆大学博士, 数据挖掘与机器学习, 博士, 数据挖掘与机器学习 at University of Chinese Academy of Sciences高中毕业, 物理+生物 组合, 高中毕业, 物理+生物 组合 at 南京师大附中
Github Skills (5)
transformers10
machine-learning10
nlp10
error-correction10
python10
Programming languages (6)
Github contributions (5)
shibing624/pycorrector
Mar 2022 - Mar 2022
pycorrector is a toolkit for text error correction. 文本纠错，实现了Kenlm，T5，MacBERT，ChatGLM3，Qwen2.5等模型应用在纠错场景，开箱即用。
Role in this project:
ML Engineer
ML EngineerContributions:2 reviews, 14 commits, 1 PR in 3 days
Contributions summary:Dian's commits focus on improving the text error correction model within the pycorrector repository. They addressed misalignment issues with UNK tokens, handled spaces and out-of-vocabulary words, and considered different scenarios during prediction. They primarily modified the `infer.py` file, which suggests they are working on the inference process and model behavior.
nlpberterror-detectionkenlmdeep-learning-models
ai-fundamental/ai-fundamental.github.io
Sep 2017 - Nov 2018
Contributions:25 commits, 23 pushes in 1 year 2 months
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.