Chiwan Park
Applied Machine Learning Engineer @ Kakao Corporation
Applied Machine Learning Engineer @ Kakao Corporation
I am an Applied Machine Learning Engineer at the AI Alignment Unit of Kakao Corporation, where I develop machine learning applications for Kakao’s various mobile and web services. My research centers on fast inference of large language models, conversational recommender systems, and machine learning on graphs. I hold an M.Sc. in Computer Science and Engineering at Seoul National University, where I researched large-scale graph processing using distributed systems under Prof. U Kang's supervision. For a comprehensive overview of my research and achievements, please refer to my full Curriculum Vitae.
Beyond my professional work, I am passionate about building data-related products. I created SolveSQL, a web-based SQL learning platform designed for data analysts. I also actively contribute to several open-source machine learning libraries and data processing frameworks including Axolotl, Liger Kernel, and Apache Flink. My personal projects are available on my GitHub Profile.
Aug. 1, 2022 - A paper, "Simple and Efficient Recommendation Strategy for Warm/Cold Sessions for RecSys Challenge 2022" is accepted to RecSys Challenge Workshop 2022.
Jun. 30, 2022 - I gave a talk named "Challenges in Real-world Recommender Systems" at KCC 2022.
Dec. 27, 2021 - I wrote a blog post named "카카오 AI추천 : 카카오의 콘텐츠 기반 필터링" to kakao Tech blog (in Korean).
Jan. 24, 2020 - A paper, "FlexGraph: Flexible partitioning and storage for scalable graph mining" is accepted to PLoS ONE.
Aug. 30, 2019 - I gave a talk named "상품 카탈로그 자동 생성 ML 모델 소개" at if (kakao)dev 2019.
M.Sc. in Computer Science and Engineering Thesis: Pre-partitioned Matrix-Vector Multiplication for Scalable Graph Mining Advisor: Prof. U Kang
B.Sc. in Earth System Sciences B.Eng. in Computer Science and Engineering (double major)
Hyunsung Lee, Sungwook Yoo, Andrew Yang, Wonjun Jang, and Chiwan Park RecSys Challenge Workshop at ACM RecSys 2022 [paper | github]
Chiwan Park, Ha-Myung Park, and U Kang PLoS ONE 15(1): e0227032 [paper | github]
Ha-Myung Park, Chiwan Park, and U Kang AAAI 2018 (demo paper) [paper | homepage]
Namyong Park, Chiwan Park, and U Kang Journal of KIISE (Vol. 43, 2016, domestic) [paper | homepage]
To be described.
As unit lead within the Kakao ART Team, I managed a research unit of over 10 members, responsible for developing and maintaining recommender systems across various Kakao services, including social networking, digital comics, e-commerce, and news platforms. Some key contributions include:
Developed a lightweight user representation model for Daum, a popular news portal with 10 million users, by leveraging topic keywords from articles to capture evolving user interests. This model served as a context input feature for bandit models, enhancing news recommendation quality.
Built personalized recommender systems for mm, an audio-only social network service like Clubhouse. I employed graph-based recommendation models, incorporating techniques like graph pruning and regularization to optimize performance.
Addressed cold-start issues for new comics on Kakao Webtoon and Piccoma by combining content-based representation learning and collaborative filtering. These recommender systems served as user-targeted marketing tools, optimizing the first conversion rate.
Created a context-aware recommender system for Kakaotalk Gift, an e-commerce feature within Kakaotalk Messenger. The system considered multiple contexts such as demographics, product popularity, and user history including clicked and purchased items, by integrating text and image embedding alongside collaborative filtering techniques.
Developed and maintained machine learning applications for shoppinghow, an eBay/Amazon-like e-commerce platform. I implemented product categorization and matching systems using Transformers and graph algorithms, applying parallel computing techniques to efficiently handle billion-scale data.
E-mail: chiwanpark at hotmail dot com
Phone: +82-10-8518-3832