Jiwon Jason Choi 최지원

This website will be deprecated in Feb. 2024

Please refer https://jiwnchoi.me (temporary URL)


Hello World! I'm Jiwon Jason Choi :)
I'm master’s student at Department of Computer Science and Engineering, Sungkyunkwan University. I'm working for SKKU IDCLab since 2021.05 advised by Jaemin Jo.
My research aims to build mixed-initiative data analytics systems, especially incorporating user intent to pipeline of automated data analysis. This includes: 1) Connect visualization with natural language using pre-trained language models. (e.g. BERT, GPT, T5); 2) Automated exploratory visual analytics which reflects user intents, data insights. I'm also interested in modern web development technology.


2024/01: Excited to start a software engineer internship at NAVER Corp!
2023/10: I’m attending VIS 2023 (Melbourne, Austrailia) in person!
2023/06: I’m attending EuroVis 2023 (Leipzig, Germany) in person to preseny my poster “Multi-Criteria Optimization for Automatic Dashboard Design”
2023/04: I’m attending PacificVis 2023 (Seoul, Korea) to present my invited poster, Also I’ll working as student volunteer!


(+82) 10-2056-9216


Visual Analytics and Information Visualization
Human-Computer (AI, Data) Interaction
Natural Language Processing and Data Science


SWE Internship
2024.01 ~ Current
LM-based automated web design
React, TypeScript, …
2021.05 ~ Current
Research Project
Automated data analysis & presentation using neural network (visualization, natural language)
Related Paper: Intentable, Waltzboard
Related Poster: Gleaner
Human-in-the-loop AutoML
Related Paper: VANAS
Robust Dimensionality Reduction
Related Paper: Project Ensemble
Teaching Assistant
2022.09 ~ Current
Human-Computer Interaction
2022 Fall, 2023 Fall
Open Source Software Practice
2022 Fall
Information Visualization
2023 Spring
Summer Bootcamp for Department of Applied AI
Open Source Software and Web Development
ROK Army AI Specialized Education
Data Visualization Basics
2023 Spring


Sungkyunkwan University
2022.09 ~ Current
Master’s Student in Computer Science and Engineering
Sungkyunkwan University
2019.03 ~ 2023.02
B.Sc. in Computer Science Education
Kyunggi Highschool
2015.03 ~ 2018.02
Peer Lecturer in School Computer Science Club


Waltzboard: Designing Instant and Interpretable Dashboards with User Intent

Preprint (Under-review)
Jiwon Choi and Jaemin Jo
We present Waltzboard, an automatic dashboard design system for exploratory data analysis, which interactively reflects the user’s analytic intent. Despite the benefit of dashboards, previous dashboard design systems often require precomputation, such as training deep-learning models, and do not adapt effectively to changes in the user’s intent during data analysis, hindering quick and flexible data exploration. To overcome these challenges, we begin by introducing our dashboard evaluation framework that quantifies the effectiveness of a dashboard design based on five key aspects: Specificity, Interestingness, Diversity, Coverage, and Parsimony. We then present a three-phase optimization algorithm designed to efficiently explore dashboard designs without the need for precomputation. Finally, we present a user interface that allows the user to dynamically specify their intent and reason for the design process. The results of our evaluation demonstrate that Waltzboard not only generates a dashboard within seconds but also supports flexible data exploration to meet diverse analytic needs.

Projection Ensemble: Visualizing the Robust Structures of Multidimensional Projections

Proceedings of Conference on 2023 IEEE Visualization & Visual Analytics (IEEE VIS), Melbourn, Australia
Myeongwon Jung, Jiwon Choi and Jaemin Jo
Paper | GitHub | Live Demo
We introduce Projection Ensemble, a novel approach for identifying and visualizing robust structures across multidimensional projections. Although multidimensional projections, such as t-Stochastic Neighbor Embedding (t-SNE), have gained popularity, their stochastic nature often leads the user to interpret the structures that arise by chance and make erroneous findings. To overcome this limitation, we present a frequent subgraph mining algorithm and a visualization interface to extract and visualize the consistent structures across multiple projections. We demonstrate that our system not only identifies trustworthy structures but also detects accidental clustering or separation of data points.

Intentable: A Mixed-Initiative System for Intent-Based Chart Captioning

Jiwon Choi and Jaemin Jo
Proceedings of Conference on 2022 IEEE Visualization & Visual Analytics (IEEE VIS), Oklahoma City, USA
We present Intentable, a mixed-initiative caption authoring system that allows the author to steer an automatic caption generation process to reflect their intents on captions. We first derive grammar for specifying the intents, i.e., a caption recipe, and build a neural network that generates caption sentences given a recipe. Our quantitative evaluation revealed that our intent-based generation system not only allows the author to engage in the generation process but also produces more fluent captions than the previous end-to-end approaches without user intervention. Finally, we demonstrate the versatility of our system such as context adaptation, unit conversion, and sentence reordering.

VANAS: A Visual Analytics System for Neural Architecture Search

Jiwon Choi, Gwon Hong, and Jaemin Jo
Proceedings of HCI KOREA 2022, Seoul, Korea
In this paper, we present VANAS, a system that analyzes and visualizes the results of the Neural Architecture Search (NAS) algorithm. First, We devised several algorithms to quantify the significance and contribution of edges, and to recommend neural architecture to users, in order to efficiently convey the vast search results. We created an overview panel to check the overall characteristics of the search space, as well as a user interface to design and analyze the neural network, based on our algorithms. We demonstrated that by analyzing NAS-Bench-101 using VANAS, users can check the importance and significance of nodes and edges that compose neural networks, and that the accuracy can be increased by adding edges recommended by VANAS.


Multi-Criteria Optimization for Automatic Dashboard Design

Jiwon Choi and Jaemin Jo
Proceedings of 2023 Eurographics Conference on Visualization (EuroVis, Poster), Leipzig, Germany
Invited to IEEE Pacific Visualization Symposium 2023, Seoul, Korea
We present Gleaner, an automatic dashboard design system that optimizes the design in terms of four design criteria, namely Specificity, Interestingness, Diversity, and Coverage. With these criteria, Gleaner not only optimizes for the expressiveness and interestingness of a single visualization but also improves the diversity and coverage of the dashboard as a whole. Users are able to express their intent for desired dashboard design to Gleaner, including specifying preferred or constrained attributes and adjusting the weight of each criterion. This flexibility in expressing intent enables Gleaner to design dashboards that are well-aligned with the user’s own analytic goals leading to more efficient data exploration.


Programming Language
Python, JavaScript / TypeScript, C, Java
MySQL(MariaDB), MongoDB
Framework and Library
React, Next.js, FastAPI, PyTorch, HuggingFace
d3.js, visx, Vega-Lite
AWS, Docker, NGiNX, Framer, Figma, Illustrator