Woondong Yeo

Jaehoon Choi

Sunwon Lee

Kyubum Lee

Seongsoon Kim

Minji Jeon

Jocelyn Barbosa

Bilal Ahmed

Hwejin Jung

Sangrak Lim

Donghyeon Kim

Jinho Lee

Buru Chang

Sungjoon Park

Bumsoo Kim

Donghyeon Park

Sunkyu Kim

Jinhyuk Lee

Won J in Yoon

Sungjun Yoon

Donghee Choi

Mujeen Sung

Mogan Gim

Hyunjae Kim

Minbyul Jeong

Quang Nguyen

Gangwoo Kim

Junhyun Lee

Cheolseong Lee

Junkyu Lee

Seongwoon Lee

Jukyoung Lee

Kwanghun Choi

Hyeokyoon Chang

Chunggyeom Kim

Heewon Lee

Wonho Shin

Byounggun Kim

Minhwan Yu

Yonggyu Park

Inyeop Lee

Nayoung Kim

Andresta Setiawan

Daehan Kim

Yookyung Koh

Doyeong Hwang

Hwisang Jeon

Miyoung Ko

Jungwoon Shin

Taeho Choi

Jaehoon Jung

Byungjoo Kim

Yoonsun Yeo

Seoyoon Kim

Jungsoo Park

Inggeol Lee

Minjae Ju

Sumin Lee

Jaehyo Yoo

Kiwoong Yoo

Jihye Kim

Sumin Seo

Chanhwi Kim

Jiwoong Sohn

The Data Mining and Information Systems (DMIS) Lab leverages AI and machine learning to address challenges in bioinformatics, drug discovery, and biomedical text mining. The lab also conducts research in natural language processing (NLP) and graph ML to refine techniques for its primary mission. Furthermore, the DMIS Lab participates in international challenges and competitions, including the DREAM Challenges and the BioASQ Challenges, contributing to collaborative efforts in biomedicine.

데이터 마이닝 및 정보 시스템(DMIS) 연구실은 AI 및 머신러닝을 활용하여 생물정보학, 신약 개발, 생물 의학 텍스트 마이닝 분야의 문제 해결에 기여한다. 주요 연구 분야는 자연어 처리(NLP) 및 그래프 ML이며, 핵심 목표 달성을 위한 기술 개선을 목표로 한다. DREAM Challenges 및 BioASQ Challenges와 같은 국제 챌린지 및 대회에 참여하여 생물 의학 분야의 공동 연구에 기여한다.

Retrosynthetic Crosstalk between Single-step Reaction and Multi-step Planning

GPO-VAE: Modeling Explainable Gene Perturbation Responses utilizing GRN-Aligned Parameter Optimization

BADGER: biologically-aware interpretable differential gene expression ranking model

CRADLE-VAE: Enhancing Single-Cell Gene Perturation Modeling with Counterfactual Reasoning-based Artifact Disentanglement

TurboHopp: Accelerated Molecule Scaffold Hopping with Consistency Models

MolPLA: A Molecular Pre-training Framework for Learning Cores, R-Groups and their Linker Joints

MulinforCPI: enhancing precision of compound–protein interaction prediction through novel perspectives on multi-level information integration

Enhancing Clinical Outcome Predictions through Auxiliary Loss and Sentence-Level Self-Attention

Evaluation of crowdsourced mortality prediction models as a framework for assessing artificial intelligence in medicine

ArkDTA: Attention Regularization guided by non-Covalent Interactions for Explainable Drug-Target Binding Affinity Prediction

Perceiver CPI: A nested cross-attention network for compound-protein interaction prediction

RecipeMind: Guiding Ingredient Choices from Food Pairing to Recipe Completion using Cascaded Set Transformer

Deep-Learning-Based Natural Language Processing of Serial Free-Text Radiological Reports for Predicting Rectal Cancer Patient Survival

Crowdsourced identification of multi-target kinase inhibitors for RET- and TAU- based disease: the Multi-Targeting Drug DREAM Challenge

RecipeBowl: A Cooking Recommender for Ingredients and Recipes Using Set Transformer

Crowdsourced mapping of unexplored target space of kinase inhibitors

Predicting mechanism of action of novel compounds using compound structure and transcriptomic signature co-embedding

The Curious Case of Analogies: Investigating Analogical Reasoning in Large Language Models

Med-PRM: Medical Reasoning Models with Stepwise, Guideline-verified Process Rewards

Assessing LLM Reasoning Steps via Principal Knowledge Grounding

Does Time Have Its Place? Temporal Heads: Where Language Models Recall Time-specific Information

Outlier-Safe Pre-Training for Robust 4-Bit Quantization of Large Language Models

Learning from Negative Samples in Biomedical Generative Entity Linking

ETHIC: Evaluating Large Language Models on Long-Context Tasks with High Information Coverage

Rationale-Guided Retrieval Augmented Generation for Medical Question Answering

Ask Optimal Questions: Aligning Large Language Models with Retriever's Preference in Conversational Search

ChroKnowledge: Unveiling Chronological Knowledge of Language Models in Multiple Domains

Monet: Mixture of Monosemantic Experts for Transformers

CompAct: Compressing Retrieved Documents Actively for Question Answering

LAPIS: Language Model-Augmented Police Investigation System

Improving Medical Reasoning through Retrieval and Self-Reflection with Retrieval-Augmented Large Language Models

CookingSense: A Culinary Knowledgebase with Multidisciplinary Assertions

Fine-tuning CLIP Text Encoders with Two-step Paraphrasing

Tree of Clarifications: Answering Ambiguous Questions with Retrieval-Augmented Large Language Models

Pre-training Intent-Aware Encoders for Zero- and Few-Shot Intent Classification

KU-DMIS-MSRA at RadSum23: Pre-trained Vision-Language Model for Radiology Report Summarization

Automatic Creation of Named Entity Recognition Datasets by Querying Phrase Representations

Optimizing Test-Time Query Representations for Dense Retrieval

LIQUID: A Framework for List Question Answering Dataset Generation

Biomedical NER for the Enterprise with Distillated BERN2 and the Kazu Framework

Saving Dense Retriever from Shortcut Dependency in Conversational Search

Generating Information-Seeking Conversations from Unlabeled Documents

Simple Questions Generate Named Entity Recognition Datasets

Consistency Training with Virtual Adversarial Discrete Perturbation

FAVIQ: FAct Verification from Information-seeking Questions

Phrase Retrieval Learns Passage Retrieval, Too

Can Language Models be Biomedical Knowledge Bases?

Simple Entity-Centric Questions Challenge Dense Retrievers

Learning Dense Representations of Phrases at Scale

Learn to Resolve Conversational Dependency: A Consistency Training Framework for Conversational Question Answering

'Killing Me' Is Not a Spolier: Spoiler Detection Model using Graph Neural Networks with Dependency Relation-Aware Attention Mechanism

Answering Questions on COVID-19 in Real-Time

Look at the First Sentence: Position Bias in Question Answering

Adversarial Subword Regularization for Robust Neural Machine Translation

Transferability of Natural Language Inference to Biomedical Question Answering

Biomedical Entity Representations with Synonym Marginalization

Contextualized Sparse Representations for Real-Time Open-Domain Question Answering

Building a PubMed knowledge graph

BioBERT: a pre-trained biomedical language representation model for biomedical text mining

Pre-trained Language Models for Biomedical Question Answering

CollaboNet: collaboration of deep neural networks for biomedical named entity recognition

A Neural Named Entity Recognition and Multi-Type Normalization Tool for Biomedical Text Mining

Can Machines Learn to Comprehend Scientific Literature?

Real-Time Open-Domain Question Answering on Wikipedia with Dense-Sparse Phrase Index

ChimerDB 4.0: an updated and expanded database of fusion genes

Ranking Paragraphs for Improving Answer Recall in Open-Domain Question Answering

Learning User Preferences and Understanding Calendar Contexts for Event Scheduling

A Deep Neural Spoiler Detection Model using a Genre-Aware Attention Mechanis
m

A Pilot Study of Biomedical Text Comprehension using an Attention-Based Deep Neural Reader: Design and Experimental Analysis

Deep learning of mutation-gene-drug relations from the literature

Drug drug interaction extraction from the literature using a recursive neural network

Chemical-gene relation extraction using recursive neural network

Name Nationality Classification with Recurrent Neural Networks

Constructing and Evaluating a Novel Crowdsourcing-based Paraphrased Opinion Spam Dataset

ChimerDB 3.0: an enhanced database for fusion genes from cancer transcriptome and literature data mining

SEMO: Searching Majority Opinions on Movies using SNS and QA Threads

BEST: Next-Generation Biomedical Entity Search Tool for Knowledge Discovery from Biomedical Literature

HiPub: translating PubMed and PMC texts to networks for knowledge discovery

BRONCO: Biomedical entity Relation ONcology COrpus for extracting gene-variant-disease-drug relations

Deep Semantic Frame-based Deceptive Opinion Spam Analysis

Smith Search: Opinion-Based Restaurant Search Engine

BOSS: context-enhanced search for biomedical objects

Consento: A New Framework for Opinion Based Entity Search and Summarization

Consento: A Consensus Search Engine for Answering Subjective Queries

BOSS: A Biomedical Object Search System

A Scalable Method for Detecting Multiple Loci Associated with Traits using TF-IDF Weighting and Association Rule Mining

Understanding and Tackling Over-Dilution in Graph Neural Networks

FlavorGraph: A large-scale food-chemical graph for generating food representations and recommending food pairings

UnionDet: Union-level Detector Towards Real-Time Human-Object Interaction Detection

Robust Neural Networks inspired by Strong Stability Preserving Rung-Kutta methods

Robust Long-Term Object Tracking via Improved Discriminative Model Prediction

Learning Graph-Based Geographical Latent Representation for Point-of-Interest Recommendation

MAPS: Multi-Agent reinforcement learning-based Portfolio management System

Content-Aware Successive Point-of-Interest Recommendation

Bipartite Link Prediction by Intra-Class Connection Based Triadic Closure

Graph Transformer Networks

KitcheNette: Prediction and Ranking Food Ingredient Pairings based on Siamese Neural Network

SAIN: Self-Attentive Integration Network for Recommendation

Self-Attention Graph Pooling

Predicting Multiple Demographic Attributes with Task Specific Embedding and Attention Network

Graph transformer networks

Self-attention graph pooling

DSigDB: drug signatures database for gene set analysis

Evaluating window joins over unbounded streams

Catching the boat with Strudel: Experiences with a web-site management system

Hotr: End-to-end human-object interaction detection with transformers

Community assessment to advance computational prediction of cancer drug combinations in a pharmacogenomic screen

On schema matching with opaque column names and data values

The Niagara internet query system

Hats: A hierarchical graph attention network for stock movement prediction

Uniondet: Union-level detector towards real-time human-object interaction detection

Content-aware hierarchical point-of-interest embedding model for successive poi recommendation.

Strudel: A web site management system

Neo-gnns: Neighborhood overlap-aware graph neural networks for link prediction

Collabonet: collaboration of deep neural networks for biomedical named entity recognition

Biomedical entity representations with synonym marginalization

BERN2: an advanced neural biomedical named entity recognition and normalization tool

Improving medical reasoning through retrieval and self-reflection with retrieval-augmented large language models

Aik Choon Tan

Donghyeon Park (박동현)

Dongwon Lee

Jeffrey Naughton

Dan Suciu

Byung-Won On

Prasenjit Mitra

Stratis Viglas

HyungJun Cho

Karen A. Ryall

Jaewoo Kang

Motivation Biomedical text mining is becoming increasingly important as the number of biomedical documents rapidly grows. With the progress in natural language processing (NLP), extracting valuable information from biomedical literature has gained popularity among researchers, and deep learning has boosted the development of effective biomedical text mining models. However, directly applying the advancements in NLP to biomedical text mining often yields unsatisfactory results due to a word distribution shift from general domain corpora to biomedical corpora. In this article, we investigate how the recently introduced pre-trained language model BERT can be adapted for biomedical corpora. Results We introduce BioBERT (Bidirectional Encoder Representations from Transformers for Biomedical Text Mining), which is a domain-specific language representation model …

Graph neural networks (GNNs) have been widely used in representation learning on graphs and achieved state-of-the-art performance in tasks such as node classification and link prediction. However, most existing GNNs are designed to learn node representations on the fixed and homogeneous graphs. The limitations especially become problematic when learning representations on a misspecified graph or a heterogeneous graph that consists of various types of nodes and edges. In this paper, we propose Graph Transformer Networks (GTNs) that are capable of generating new graph structures, which involve identifying useful connections between unconnected nodes on the original graph, while learning effective node representation on the new graphs in an end-to-end fashion. Graph Transformer layer, a core layer of GTNs, learns a soft selection of edge types and composite relations for generating useful multi-hop connections so-call meta-paths. Our experiments show that GTNs learn new graph structures, based on data and tasks without domain knowledge, and yield powerful node representation via convolution on the new graphs. Without domain-specific graph preprocessing, GTNs achieved the best performance in all three benchmark node classification tasks against the state-of-the-art methods that require pre-defined meta-paths from domain knowledge.

Advanced methods of applying deep learning to structured data such as graphs have been proposed in recent years. In particular, studies have focused on generalizing convolutional neural networks to graph data, which includes redefining the convolution and the downsampling (pooling) operations for graphs. The method of generalizing the convolution operation to graphs has been proven to improve performance and is widely used. However, the method of applying downsampling to graphs is still difficult to perform and has room for improvement. In this paper, we propose a graph pooling method based on self-attention. Self-attention using graph convolution allows our pooling method to consider both node features and graph topology. To ensure a fair comparison, the same training procedures and model architectures were used for the existing pooling methods and our method. The experimental results demonstrate that our method achieves superior graph classification performance on the benchmark datasets using a reasonable number of parameters.

Summary: We report the creation of Drug Signatures Database (DSigDB), a new gene set resource that relates drugs/compounds and their target genes, for gene set enrichment analysis (GSEA). DSigDB currently holds 22 527 gene sets, consists of 17 389 unique compounds covering 19 531 genes. We also developed an online DSigDB resource that allows users to search, view and download drugs/compounds and gene sets. DSigDB gene sets provide seamless integration to GSEA software for linking gene expressions with drugs/compounds for drug repurposing and translational research. Availability and implementation: DSigDB is freely available for non-commercial use at http://tanlab.ucdenver.edu/DSigDB. Supplementary information: Supplementary data are available at Bioinformatics online. Contact: aikchoon.tan@ucdenver.edu

We investigate algorithms for evaluating sliding window joins over pairs of unbounded streams. We introduce a unittime-basis cost model to analyze the expected performance of these algorithms. Using this cost model, we propose strategies for maximizing the efficiency of processing joins in three scenarios. First, we consider the case where one stream is much faster than the other. We show that asymmetric combinations of join algorithms,(eg, hash join on one input, nested-loops join on the other) can outperform symmetric join algorithm implementations. Second, we investigate the case where system resources are insufficient to keep up with the input streams. We show that we can maximize the number of join result tuples produced in this case by properly allocating computing resources across the two input streams. Finally, we investigate strategies for maximizing the number of result tuples produced when memory is limited, and show that proper memory allocation across the two input streams can result in significantly lower resource usage and/or more result tuples produced.

The Strudel system applies concepts from database management systems to the process of building Web sites. Strudel's key idea is separating the management of the site's data, the creation and management of the site's structure, and the visual presentation of the site's pages. First, the site builder creates a uniform model of all data available at the site. Second, the builder uses this model to declaratively define the Web site's structure by applying a “site-definition query” to the underlying data. The result of evaluating this query is a “site graph”, which represents both the site's content and structure. Third, the builder specifies the visual presentation of pages in Strudel's HTML-template language. The data model underlying Strudel is a semi-structured model of labeled directed graphs. We describe Strudel's key characteristics, report on our experiences using Strudel, and present the technical problems that arose from …

Professor

DMIS Lab

교수님 소개

강재우

연구실 소개

연구실 키워드

알럼나이가 간 회사

연구실 지원 방법

대우 조건

연도별 논문 인용수

저널 & 학회 논문 출판 수

연구실 논문 목록

The Curious Case of Analogies: Investigating Analogical Reasoning in Large Language Models

Retrosynthetic Crosstalk between Single-step Reaction and Multi-step Planning

GPO-VAE: Modeling Explainable Gene Perturbation Responses utilizing GRN-Aligned Parameter Optimization

연구실 보유 특허

대학별 입시끈

우리 연구실 모집

교수님 소개

강재우

연구실 키워드

알럼나이가 간 회사

연구실 지원 방법

대우 조건

연도별 논문 인용수

저널 & 학회 논문 출판 수

연구실 논문 목록

The Curious Case of Analogies: Investigating Analogical Reasoning in Large Language Models

Retrosynthetic Crosstalk between Single-step Reaction and Multi-step Planning

GPO-VAE: Modeling Explainable Gene Perturbation Responses utilizing GRN-Aligned Parameter Optimization

연구실 보유 특허

최근 3년간 연구실적 높은 연구실

대학별 입시끈

우리 연구실 모집