About Me

I am a Ph.D. student advised by Prof. Linglong Kong and Prof. Bei Jiang at the University of Alberta (UofA). I obtained my M.S. degree at the University of Minnesota with Prof. Yang Li and Prof. Haiyang Wang and B.S. degree at Beijing University of Posts and Telecommunications(BUPT).

I am especially passionate about research in Statistics, Machine learning, Deep learning, and Natural language processing and their application and impact on society. As such, I work at the intersection of applying statistical machine learning methods and natural language processing algorithms to investigate social impact.

Interview:

University of Alberta Folio

Academic Research and Conference Papers:

From physical space to cyberspace: Recessive gender biases in social media mirror the real world (Cities)

• We propose a new Framework to analyze and quantify recessive gender biases in social media, using Hong Kong’s Twitter data as a case study.
• Our framework demonstrates the temporal trends and spatial distribution of these biases, revealing that the gender biases in Hong Kong’s virtual spaces closely mirror those in the physical world, highlighting the persistent influence of gender stereotypes across both realms.

Debiasing with Sufficient Projection: A General Theoretical Framework for Vector Representations (NAACL 2024)

• We propose a novel framework to reduce bias by transforming vector representations to an unbiased subspace using sufficient projection.

Gaussian differential privacy on Riemannian manifolds (NeurIPS 2023)

• We develop an advanced approach for extending Gaussian Differential Privacy (GDP) to general Riemannian manifolds.

Local Differential Privacy for Population Quantile Estimation (ICML 2023)

• Developed a novel approach for estimating population quantiles with Local Differential Privacy.

Quantile Fairness Regression with Conformalized Prediction Intervals (NeurIPS 2022)

• Presented a pioneering study on quantile fairness regression at NeurIPS 2022.
• Proposed a novel conformalized prediction interval to assess fairness algorithm uncertainty and provide fair prediction intervals.

Reducing Gender Bias in GloVe Word Embeddings Using Causal Inference (AAAI 2022)

• We propose a method for gender bias in GloVe word embeddings while retaining semantics information.

AI-driven Combat Against Bias in Job Recruitment (Canada-UK AI Initiative)

• Led an international project to identify and mitigate gender and ethnic bias in the job market using AI.
• Managed a diverse team of postdocs and Ph.D. students from Canada and the UK.
• Collaboration with sociology teams for interdisciplinary work.
• Published outcomes in reputable journals and conferences.

Synthetic Health Data Generation with Deep Learning (Replica Analytics)

• Internship at Replica Analytics, working on synthesizing structured health data.
• Developed state-of-the-art deep learning models to generate longitudinal synthetic data.
• Contributed to critical modeling discussions and utility evaluation methods.
• Research paper accepted at BMC Medical Research Methodology 2023.

Predicting Carbon Percentage in Alberta Soil (Alberta Biodiversity Monitoring Institute)

• Collaborated on an interdisciplinary project for predicting carbon soil percentage.
• Developed a novel screening variable method with bootstraps group lasso for analysis.
• Submitted research paper to Science of the Total Environment in 2021.

Reducing Selection Bias in Counterfactual Reasoning (NeurIPS 2019 Workshop)

• Contributed to a research project addressing selection bias in counterfactual reasoning.
• Proposed a novel synthetic data-generating mechanism.
• Paper presented at NeurIPS 2019 Workshop on “Do the right thing.”

Cloud User Interest Analysis in P2P Systems (IEEE INFOCOM 2020)

• Conducted measurements and analysis of cloud peers’ interest in P2P systems.
• Built web scrapers and performed data analysis and visualization.
• Utilized machine learning (clustering algorithm) to gain insights.

Invited Talk:

Ding, L, (Author & Presenter), Society for Industrial and Applied Mathematics (SIAM) Spring 2023, “Word embeddings via causal inference: Gender bias reducing and semantic information preserving” the University of Texas at Arlington, Arlington, TX. (May 2nd, 2023)

Internships and Personal Projects:

1. Augmented Decision-making for Innovation Management (Human Resources Company)

• Led a project that utilized various language modeling methods to cluster ideas and comments efficiently.
• Implemented NLP models including TF-IDF, averaging word vectors, doc2vec, and RNN with BERT embeddings.
• Successfully improved innovation management and resource allocation.

1. Medical Insurance Cost Prediction using Recurrent Neural Networks

• Internship at Guangzhou Huazi Software Technology Co in 2018.
• Built RNN models to predict medical insurance costs for government authorities.
• Worked on NLP word correction for OCR with Kneser-Ney smoothing and noisy channel model.

2. Named Entity Recognition with RNN and Tensorflow (2018)

• Developed a bi-directional LSTM model with character embeddings for Named Entity Recognition.
• Utilized Glove word embeddings and conditional random field for accurate predictions.

4. Neural-based Dependency Parser Implementation (2017)

• Implemented a neural-based dependency parser using fully connected dense layers.
• Incorporated word embeddings and POS tag features for improved accuracy.

5. Cooperation of Clustering and Differential Evolution Algorithm (2015)

• Explored the combination of different clustering algorithms with Differential Evolution.
• Improved searching ability and convergence speed on challenging optimization problems.

6. 3D Human Faces Modeling at SAMSUNG Advanced Institute of Technology (2015)

• Internship focused on 3D human faces modeling.
• Developed Matlab and C++ programs for face point cloud data processing and ECG signals analysis.