NLP & Rights Metadata
A NER + LLM “division of labor” pipeline for Korean openly licensed works, with per-field provenance — built in the KOCCA consortium with 무하유, HM컴퍼니, 한국저작권위원회, and 한국문화정보원.
67-field metadata schema · 144k-work KOGL corpus
Latest Two papers accepted at IC-EEECS 2026
We build systems that understand data. 우리는 데이터를 이해하는 시스템을 만듭니다.
The Database Lab at Soongsil University — two decades of data systems research, from the FAST flash translation layer to on-device Korean voice-phishing detection and trustworthy rights-metadata AI, led by Prof. Dong-Joo Park.
712
citations · FAST FTL (ACM TECS 2007)
19
alumni · 2 Ph.D. + 17 M.S.
60
publications · 2001 – 2026
2
patents · 1 granted + 1 filed
9,132
KOCW views · Database open courseware
Research
Research Areas연구 분야
Four lines of work, one thread: systems that make data fast, recoverable, and trustworthy.
Voice Phishing Detection
KorCCVi — Korea’s first labeled vishing corpus — and a model line from KoBERT to multimodal audio–text fusion and privacy-preserving real-time detection.
KorCCVi — the first labeled Korean voice-phishing dataset
Digital Forensics
Recovery of deleted and wiped files on NAND flash, SSD TRIM selective recovery, YAFFS2 metadata, and SQLite deleted-record and messenger forensics.
10+ M.S. theses on flash & SQLite recovery (2010–2020)
Flash Storage & FTL
Log-buffer flash translation layers (FAST, LSTAFF*), hybrid address mapping, flash-aware B-tree indexes, and crash recovery for NAND flash and SSDs.
FAST FTL — 712 citations (ACM TECS 2007)
News
News소식
Jul 1, 2026
paper Two papers accepted at IC-EEECS 2026 — the NER–LLM consolidation pipeline for metadata extraction from Korean public-domain rights documents, and the KLUE-BERT B-I-O training-data composition analysis.
Jun 8, 2026
lab Recruiting graduate students — Soongsil 2026 fall (후기) graduate admissions: applications June 8–16, interviews June 26, results July 21 (grad.ssu.ac.kr).
Jun 1, 2026
paper Research impact: the FAST flash translation layer paper (ACM TECS 2007) has passed 700 citations.
May 1, 2026
paper Revised manuscript on lightweight on-device vishing detection submitted to IEEE Access (revise-and-resubmit).
Selected Work
Selected Publications대표 논문
Six papers that trace the lab’s arc — from arbitrated NER–LLM metadata extraction back to the FAST FTL.
Arbitrating Two Extractors: A NER-LLM Consolidation Pipeline for Metadata Extraction from Korean Public Domain Right Documents
Milandu Keith Moussavou Boussougou, SeongHun Hwang, Dong-Joo Park
IC-EEECS 2026 (short paper, to appear)
Enhancing Voice Phishing Detection Using Multilingual Back-Translation and SMOTE: An Empirical Study
Milandu Keith Moussavou Boussougou, Prince Hamandawana, Dong-Joo Park
IEEE Access 13:37946–37965
⚠ under verification DOI
Attention-Based 1D CNN-BiLSTM Hybrid Model Enhanced with FastText Word Embedding for Korean Voice Phishing Detection
Milandu Keith Moussavou Boussougou, Dong-Joo Park
Mathematics (MDPI) 11(14):3217
한국어 언어 모델을 활용한 보이스피싱 탐지 기능 개선
Milandu Keith Moussavou Boussougou, Dong-Joo Park
KIPS Transactions on Software and Data Engineering 11(10):437–446
A survey of Flash Translation Layer
Tae-Sun Chung, Dong-Joo Park, Sangwon Park, Dong-Ho Lee, Sang-Won Lee, Ha-Joo Song
Journal of Systems Architecture 55(5-6):332–343
267 citations DOI
A log buffer-based flash translation layer using fully-associative sector translation
Sang-Won Lee, Dong-Joo Park, Tae-Sun Chung, Dong-Ho Lee, Sangwon Park, Ha-Joo Song
ACM Transactions on Embedded Computing Systems 6(3), Article 18
712 citations DOI
Join · 모집
Join the Database Lab
We are recruiting M.S./Ph.D. students and undergraduate interns. Work on funded national projects — voice-phishing detection, rights-metadata AI, flash storage — publish internationally, and build systems that ship.