Demo Data from SIGIR 2026 / reviewers in the wild / expert
Jinkwon Lee
dblp:424/4522
· DBLP profile ↗
4ranked-venue papers
4as first author
4since 2021 · last 2026
0009-0001-8301-500XORCID · corroborated
Domains — the database's venue-derived domains; a paper can count in several
Databases, data management, data science and information retrieval · 3 · 3 first-author · 3 since 2021Artificial intelligence and machine learning · 1 · 1 first-author · 1 since 2021Computer networks · 1 · 1 first-author · 1 since 2021
Expertise — from the expertise taxonomy: the topics of the expert's papers under the CCF categories. A weight counts papers with recency: 1 for a paper about the topic, 0.3 when the topic is its context, halved every five years.
| Databases, data mining, and information retrieval
2 papers |
Recommender systems · 77% Web and social media mining · 23% | |
| Computer graphics and multimedia
1 paper |
Multimedia analysis and retrieval · 100% | |
| Artificial intelligence
1 paper |
Efficient and distributed learning · 100% |
Topics — the 3 heaviest of 5, each with the papers that count most for it
| Topic | Weight | Papers | Last | Evidence papers |
|---|---|---|---|---|
Recommender systems
explainable recommendation |
1.0 | 1 | 2026 | PEARL: Profile-based Explainable Agent for Edge LLM Recommendation via Latent Decomposition · SIGIR 2026 |
Machine learning › Efficient and distributed learning
edge computing |
0.3 | 1 | 2026 | PEARL: Profile-based Explainable Agent for Edge LLM Recommendation via Latent Decomposition · SIGIR 2026 |
Machine learning › Efficient and distributed learning
on-device inference |
0.3 | 1 | 2026 | PEARL: Profile-based Explainable Agent for Edge LLM Recommendation via Latent Decomposition · SIGIR 2026 |
Methods — techniques the papers use, named apart from their topics
profile alignment · 2.0perturbation analysis · 2.0multimodal retrieval-augmented fine-tuning · 2.0latent decomposition · 2.0
| Year | Publication | Venue | Position |
|---|---|---|---|
| 2026 | Explainable Cognitive Task Classification in Pediatric EEG Using CPCC-Based Functional Connectivity Images
Jinkwon Lee, Seohyeon Hong, Hayoung Oh 0002 |
PAKDD (3) | 1 |
| 2026 | PEARL: Profile-based Explainable Agent for Edge LLM Recommendation via Latent DecompositionabstractWhich on-device LLM best fits a given user? 2B-class edge models exhibit user-specific behavioral variation: one may better align with genre-sensitive recommendations, another with procedural instructions, and no single model dominates across all users or domains. We present PEARL (Profile-based Explainable Agent for edge-model Recommendation via Latent Decomposition), which selects the best-fitting 2B-class edge LLM from a structured user profile and generates a natural-language explanation grounded in interpretable latent alignment dimensions. Its core, Explainable Latent Decomposition (ELD), maps per-model profile-alignment fingerprints into a shared low-correlation latent subspace for direct profile-to-model matching. On PersonaLens (200 profiles, 8,521/1,974 TSD/TMD dialogues), PEARL achieves P = 2.23 \pm .02 on TSD, outperforming the best single-fixed edge model by +0.11 and closing 35.0% of the Oracle-UB gap; on TMD, PEARL achieves P = 2.14 \pm .03 (+0.11, 34.0% Oracle-UB gap closure). A perturbation analysis shows that masking the top-1 ELD dimension alters 60% of recommendations (Δ P = -0.43), with a modest 1.3× specificity ratio over a random-dimension control. Jinkwon Lee, Hayoung Oh 0002 |
SIGIR | 1 |
| 2026 | Fine-Tuned LLMs for Flow-Based Intrusion Detection in Smart Agriculture via Semantic Augmentation
Jinkwon Lee, Youn Jun Seong, Hayoung Oh 0002 |
WoWMoM | 1 |
| 2026 | MerFT: A Framework for Social Conflict Meme Exploration via Multimodal Retrieval-Augmented Fine-tuning
Jinkwon Lee, Giseong Kim, HaeJi Yang, Dongyoung Tcha, Hayoung Oh 0002 |
WSDM | 1 |