Fashion Account Discovery
KDD 2017 ML4Fashion Workshop
Classification and crawling Twitter data to identify fashion accounts in social networks
Doris Jung-Lin Lee
I am a second-year graduate student at University of Illinois, Urbana-Champaign. My current research interest lies in database, data mining and human-computer interaction. I am currently working on research problems related to interactive visual data exploration systems.
I did my undergrad in UC Berkeley where I graduated with a double major in physics and astrophysics. During undergrad, I worked on crowdsourcing, genomics visualizations, fabrication, astrophysical simulations, and data mining for energy sciences and cosmology for various periods of time. I am interested in creating technology that bridges the knowledge and skillsets of different communities.
Publications
Crowdclass: Designing classification-based citizen science learning modulesHCOMP 2016[Project Page] [Code] [PDF] [Slides] [Supplementary Material] |
Skintillates: Design and Prototype Epidermal InteractionsDIS 2016 Honorable Mention Award[Project Page] [PDF] |
‘I dont want to wear a screen’: Probing perceptions of and possibilities for dynamic displays on clothingCHI 2016 Best Paper Award[Project Page] [PDF] |
Past Projects
Social Behavioural Data Mining
Energy Usage Patterns
AAAI 2017
AI for Smart Grids & Buildings Workshop
Comparison of Clustering Techniques for Residential Energy Behavior using Smart Meter Data
Fabrication tools & Wearables
Shrinky Circuits
Developing a rapid prototyping technique for building circuit boards without involving conventional PCB/chemical etching procedures.
Project Jacquard
Collaboration with Google Advanced Technologies and Projects (ATAP) group on a new type of interactive wearable technology.
Skintillates
Creating low-cost, accessible fabrication technique for on-skin wearable electronics that can integrate with a variety of electronic components.
Cosmological Data Mining
Pattern Discovery and Large-Scale Data Mining on Cosmological Datasets. [Poster]Workshop on Algorithms for Modern Massive Data Sets (MMDS 2016).
RC3 Mosaics
Designing a pipeline for generating scientifically calibrated images of large nearby RC3 galaxies using an adaptive algorithm for positional update.
Halo Finder
Applying unsupervised machine learning algorithms to cosmological simulations for finding dark matter haloes.
SDSS Imaging Systematics
Investigating how systematics affect the imaging data quality from the Sloan Digital Sky Survey.
Astrophysical Fluid Simulations
Star Formation
Magnetohydrodynamics, adaptive mesh refinement simulations for the evolution of a collapsing dense core.
Accretion Disk
Investigating the effects of Papaloizou-Pringle and magnetorotational instabilities in accretion disk torus.
astroSim-tutorial
Jupyter Notebook tutorials on how to build, run, and analyze astrophysical simulations.