I lead a team of Applied Scientists working on challenging problems in Amazon Search Ad using AI/ML/NLP and data science. Prior to joining Amazon, I was a senior direct of research at Yahoo Research, heading the Data Science Research team and working on natural language process, machine learning and visualization. Before joining Yahoo I was at AT&T Labs, working on information visualization and recommender systems. From 2001 to 2007, I worked at Wolfram Research, developing numerical and graph theoritical software. Prior to April 2001, I worked in the at Daresbury Laboratory, conducting research in algorithms for high performance computing.
Dr. Yifan Hu (胡一凡)
Seattle, Washington


SmartGD: A GAN-Based Graph Drawing Framework for Diverse Aesthetic Goals (IEEE TVCG, 2023)
Embeddings of the BERT word-piece tokens
What's in a name? Gender Classification of Names with Character Based ML Models (ECML-PKDD 2021)
Two papers, one on hatespeech detection and another on language modeling, were accepted for EMNLP 2021.
Our paper, Collaborative Filtering for Implicit Feedback Datasets , wins the 2017 IEEE ICDM 10-Year Highest-Impact Paper Award!

There is More to Streamgraphs than Movies (demo. Eurovis 2016 best paper)
Coloring edges/countries to disambiguate graph and map drawings

Real time, topical clustering and visualization of the Twitter stream (the website went down with my move from AT&T, so paper only unfortunately).
A Maxent-Stress Model for Graph Layout
Visualizing the tree of life
A Gallery of Large Graphs featured by: IEEE Spectrum
Bundling edges fast with the agglomerative multilevel bundling algorithm .

Visualizing dynamic info with Trend Map

new The landscape of Music, Movie, TV, Books and the Internet

new Visualizing Relations and Structures as Maps: GMap

new Visualizing the Mathematics Genealogy Trees.