ICDM 2013

119 papers
The 2013 IEEE International Conference on Data Mining (ICDM) featured 162 papers addressing diverse challenges in data science and pattern recognition. Research themes focused heavily on social network analysis, including social influence modeling, and the extraction of spatio-temporal patterns from uncertain databases. Data stream processing also remained a core topic, with work on adaptive model trees for real-time analysis. Technical contributions emphasized high-dimensional data management through Nonnegative Matrix Factorization (NMF) and the development of distributed machine learning frameworks like the MLI API. Other notable research directions included multi-label learning, supervised label dependency propagation, and strategies for handling imbalanced multi-class data. Prolific authors contributing to this year's proceedings included Philip S. Yu, Enhong Chen, Thomas Seidl, and Jiawei Han. Their work, alongside others, explored practical applications such as predicting consumer response spikes and improving the efficiency of matrix completion in high dimensions.

Top Research Directions

Research directions most represented by ICDM 2013 papers.

Distributed Submodular Maximization 1
Interpretable Matrix Decomposition via Column Subset Selection 1
Sparse Random Projections for Subspace Embeddings 1
Incremental Matrix Inversion via Rank One Updates 1
Randomized Sketching for Scalable Matrix Computations 1
Predicting Missing Links in Incomplete Knowledge Graphs 1
The Geometric Curse of Dimensionality 1
Latent Space Alignment for Cold Start Recommendation 1
Dynamic Preference Modeling in Interactive Systems 1
The Scalability Bottleneck of Singular Value Decomposition 1

Top Authors

AuthorPapers
Philip S. Yu5
Aidong Zhang4
Chengqi Zhang4
Jing Gao3
Xingquan Zhu3
Chi Wang2
Christian Böhm2
Claudia Plant2
Ee-Peng Lim2
Enhong Chen2
Francesco Bonchi2
Hiroshi Sawada2
Jiawei Han2
Ling Chen2
Longbing Cao2
Ming-syan Chen2
Nan Du2
Ping Luo2
Ryota Tomioka2
Shou-De Lin2

Impactful Papers

Papers from ICDM 2013 that shaped research clusters.

Foundational

Influential

Papers

Title
Regularization Paths for Sparse Nonnegative Least Squares Problems with Applications to Life Cycle Assessment Tree Discovery
Sampling Heterogeneous Networks
Search Behavior Based Latent Semantic User Segmentation for Advertising Targeting
Semantic Frame-Based Document Representation for Comparable Corpora
Sparse K-Means with the l_q(0leq q< 1) Constraint for High-Dimensional Data Clustering
Spatio-Temporal Topic Modeling in Mobile Social Media for Location Recommendation
Spectral Subspace Clustering for Graphs with Feature Vectors
Statistical Inference of Protein "LEGO Bricks"
Stochastic Blockmodel with Cluster Overlap, Relevance Selection, and Similarity-Based Smoothing
Subgraph Enumeration in Dynamic Graphs
Tag-Weighted Dirichlet Allocation
The Pairwise Gaussian Random Field for High-Dimensional Data Imputation
TopicSketch: Real-Time Bursty Topic Detection from Twitter
Transfer Learning across Networks for Collective Classification.
Tree-Like Structure in Large Social and Information Networks
UBLF: An Upper Bound Based Approach to Discover Influential Nodes in Social Networks
Utilizing URLs Position to Estimate Intrinsic Query-URL Relevance
Validating Network Value of Influencers by Means of Explanations
Weighted-Object Ensemble Clustering

« prev Page 3/3