ACM SIGKDD is pleased to announce the winners of the best paper awards from the set of papers submitted to the research track of the KDD 2014 conference.
The overall best research track paper award goes to the paper:
Reducing the Sampling Complexity of Topic Models
Aaron Q Li, Carnegie Mellon University; Amr Ahmed, Google Inc.; Sujith Ravi, Google Inc.; Alexander J Smola, Carnegie Mellon University and Google Inc.
This paper presents an approximate sampler for topic models that theoretically and experimentally outperforms existing samplers thereby allowing topic models to scale to industry-scale datasets.
The runner up to the best research track paper is:
Efficient SimRank Computation via Linearization
Takanori Maehara, National Institute of Informatics; Mitsuru Kusumoto, Preferred Infrastructure, Inc; Ken-ichi Kawarabayashi, National Institute of Informatics.
This paper presents a new method of computing SimRank so as to substantially reduce its memory requirements, thereby making SimRank practical on very large graphs.
The best student paper award goes to:
An Efficient Algorithm For Weak Hierarchical Lasso
Yashu Liu, Arizona State University; Jie Wang, Arizona State University; Jieping Ye, Arizona State University;
The hierarchical lasso is an effective way of regularizing parameters of models that attempt to capture non-linear feature interaction. This paper presents algorithms for tackling the non-convexity that arises in using such regularizers.
The runner up for the best student paper award is:
Fast Flux Discriminant for Large-Scale Sparse Nonlinear Classification
Wenlin Chen, Washington University in St. Louis; Yixin Chen, Washington University in St. Louis; Kilian Q. Weinberger, Washington University in St. Louis;
This paper presents a smart method of transforming features prior to log-linear classification so as to capture non-linear interactions without sacrificing scalability and interpretability.
We thank the best paper award committee comprising of Deepak Agarwal, Charles Elkan, Thorsten Joachims, Eamonn Keogh, Kristian Kersting, Sunita Sarawagi, and the anonymous reviewers for selecting the award papers.