Advances In Knowledge Discovery And Data Mining, Part I: 14th Pacific-asia Conference, Pakdd 2010, Hyderabat, India, June 21-24, 2010, Proceedings (lecture Notes In Computer Science)

Cover......Page 1 Advances in Knowledge Discovery and Data Mining, Part I......Page 3 Lecture Notes in Artificial Intelligence 6118......Page 2 ISBN-10 3642136567......Page 4 Preface......Page 5 PAKDD 2010 Conference Organization......Page 7 Table of Contents – Part I......Page 13 Table of Contents – Part II......Page 18 Empower People with Knowledge: The Next Frontier for Web Search......Page 23 Discovery of Patterns in Global Earth Science Data Using Data Mining......Page 24 Game Theoretic Approaches to Knowledge Discovery and Data Mining......Page 25 Introduction......Page 26 Shared Neighbor Similarity and the RSC Model......Page 28 A Hill-Climbing Heuristic......Page 29 Complexity Analysis......Page 30 Experimental Results......Page 32 Biological Data......Page 33 Image Data......Page 34 Categorical Data......Page 35 References......Page 36 Introduction......Page 38 VAT......Page 39 Improved VAT (iVAT)......Page 40 Automated VAT (aVAT)......Page 42 Test Datasets......Page 43 Results and Analysis......Page 44 Algorithm Comparison......Page 46 Conclusion......Page 47 References......Page 48 Introduction......Page 50 Preliminaries......Page 52 Algorithm Description......Page 53 Synthetic Datasets......Page 56 Real-World Datasets......Page 57 References......Page 59 Introduction......Page 60 Related Work......Page 61 Coding Categorical Data......Page 62 Coding Numerical Data......Page 63 A Coding Scheme for Integrative Clustering......Page 64 Automatically Selecting the Number of Clusters k......Page 65 Synthetic Data......Page 66 Real Data......Page 67 Finding the Optimal k......Page 68 References......Page 69 Introduction......Page 70 Sum Squared Residue......Page 71 Data Transformations......Page 72 Column Standard Deviation Normalization (SDN)......Page 73 Experimental Results......Page 74 Conclusion and Remark......Page 76 References......Page 77 Introduction......Page 78 Problem of Matrix Approximation Optimization......Page 80 Problem Definition......Page 81 The Proposed Algorithms......Page 82 Determining Initial Centroids......Page 83 Communities Discovering......Page 84 Performance Evaluation......Page 85 Results and Discussions......Page 86 References......Page 89 Introduction......Page 90 Related Work......Page 91 Antagonistic Group......Page 92 A-Group Mining Algorithm......Page 94 Performance and Case Studies......Page 98 References......Page 101 Introduction......Page 103 Related Work......Page 104 Framework for Temporal Analysis......Page 105 Experiments......Page 107 References......Page 112 Introduction and Motivation......Page 113 Background and Related Work......Page 114 Proposed Method......Page 115 Experimental Setup......Page 116 Experimental Results......Page 117 Conclusion......Page 119 References......Page 120 Introduction......Page 121 Background......Page 122 Estimate on Expectation......Page 123 GEE Algorithm......Page 124 Experiments......Page 125 Discussion......Page 127 References......Page 128 Introduction......Page 129 Related Work......Page 130 Receiver Operating Characteristics Curves......Page 131 Our Method......Page 132 Experimental Design and Data Sets......Page 138 Results and Discussion......Page 139 Conclusion......Page 141 References......Page 142 Introduction......Page 143 Related Work......Page 144 Preliminaries......Page 145 PSM......Page 146 rPCL and ePCL......Page 147 Experiment Setup......Page 150 Parameters Setting......Page 151 Efficiency......Page 152 References......Page 154 Introduction......Page 156 Core Vector Machine with Minimum Enclosing Ball......Page 157 Justification of the Rough Margin......Page 158 Rough Margin Based Core Vector Machine......Page 159 Solving Rough Margin Based CVM......Page 160 Experiments......Page 161 References......Page 163 Introduction......Page 164 Feature Relevance......Page 166 Details of the Algorithm......Page 167 Experimental Results......Page 169 References......Page 171 Introduction......Page 172 Classification Using Emerging Patterns......Page 173 Crisp Emerging Pattern Mining (CEPM)......Page 174 Experimental Results......Page 176 Conclusions......Page 178 References......Page 179 Introduction......Page 180 Related Work......Page 181 Background and Problem Statement......Page 182 Multidimensional Local Recoding......Page 183 Metric for Multidimensional Local Recoding......Page 184 Algorithm for Hiding Emerging Patterns......Page 186 Experimental Evaluation......Page 188 References......Page 191 Introduction......Page 193 Privacy and Anonymization Model......Page 195 Integrating Generalization and Suppression......Page 196 Finding a Good Suppression Scenario for a Cut......Page 197 Addressing Efficiency and Scalability Issues......Page 198 A Multi-round Approach......Page 199 Information Loss Evaluation......Page 200 Conclusion......Page 201 References......Page 202 Introduction......Page 203 Motivation......Page 204 $(k,\epsilon,l)$-Anonymity......Page 205 The Algorithm......Page 207 Experiments......Page 208 Conclusion......Page 209 References......Page 210 Introduction......Page 211 Latent Dirichlet Allocation......Page 212 Protocol......Page 215 Implementations and Performance......Page 217 Experiments......Page 218 References......Page 219 Introduction......Page 220 Secure Network Aggregation......Page 221 Preliminaries......Page 222 Private Product Correlation Protocol......Page 223 Efficiency Improvements......Page 226 Performance......Page 227 Conclusion......Page 228 References......Page 229 Introduction......Page 230 Multivariate Data Swapping......Page 231 Multivariate EDS......Page 232 Multivariate EWS......Page 233 Privacy Analysis......Page 234 Privacy on Binormal Distribution......Page 235 Comparison of Data Utility......Page 236 References......Page 237 Introduction......Page 238 Correspondence Analysis Framework......Page 239 Representative-Based Correspondence Clustering Algorithms......Page 240 Agreement Assessment by Forward and Backward Re-clustering Techniques and Co-occurrence Matrices......Page 242 Experiment Investigating Variance Reduction......Page 244 Experiment for Representative-Based Correspondence Clustering with Different Methods of Initial Representative Settings......Page 246 Related Work......Page 247 References......Page 248 Introduction......Page 250 Related Work......Page 251 Computing Fréchet Distance......Page 252 Segmentation......Page 254 Inter-grid Concatenation......Page 255 Experiments......Page 256 References......Page 258 Introduction......Page 260 Related Work......Page 261 Subseries Join......Page 262 Experimental Evaluations......Page 263 References......Page 267 Introduction......Page 268 Tensor Tools......Page 269 TWaveCluster......Page 271 Dataset......Page 272 Clustering......Page 273 Conclusions......Page 274 References......Page 275 Introduction......Page 276 Classical PSO......Page 277 Coordinate PSO with Dynamic Piecewise Linear Chaotic Map and Dynamic Nonlinear Inertia Weights......Page 278 Motivating Concepts......Page 279 Spatial Clustering with Obstacles Constraints Based on PNPSO and Improved K-Medoids......Page 280 Results and Discussion......Page 281 Conclusions......Page 282 References......Page 283 Introduction......Page 284 Related Work......Page 285 Definitions......Page 286 GA-Based Negative Sequential Pattern Mining Algorithm......Page 287 Population and Selection......Page 288 Fitness Function......Page 289 Algorithm Description......Page 290 Experiments......Page 291 References......Page 295 Introduction......Page 296 Background......Page 297 The Weighted Association Rule Mining (WARM) Problem......Page 298 Valency Model......Page 299 Rule Evaluation Methodology......Page 302 Principal Components Analysis of the Rule Bases......Page 303 Case Study Zoo Dataset......Page 305 References......Page 306 Motivation......Page 308 Overview of the Method......Page 309 Related Work and Contributions......Page 310 Sequential Pattern Mining......Page 311 Significance of Subsequence Patterns......Page 312 Generative Process......Page 313 Relative Frequency......Page 314 Itemset-Wise Reference Model......Page 316 Experiments......Page 317 From Clusters to Itemset-Sequences......Page 318 Significant Patterns......Page 319 References......Page 321 Introduction......Page 322 Related Work......Page 323 Problem Setting......Page 324 Improved Interesting Itemsets Algorithm......Page 325 Association Rules Algorithm......Page 327 Experiments......Page 328 References......Page 331 Introduction......Page 332 Problem Definition......Page 333 Mining Closed Episodes......Page 334 Experiments......Page 337 Conclusion......Page 339 References......Page 340 Motivation......Page 341 Definition and Properties......Page 343 All-Pair Refined Local Maxima Search (ARLM)......Page 344 Approximate Greedy Maximum Maxima Search (AGMM)......Page 345 Experiments......Page 346 Synthetic Datasets......Page 347 References......Page 349 Introduction......Page 350 Related Work......Page 351 Overview......Page 352 Methods......Page 354 Evaluation......Page 357 References......Page 360 Introduction......Page 362 Related Work......Page 363 Using Association Rules to Expand User Profiles......Page 364 Evaluation Metrics and Datset......Page 365 Experiment Results......Page 366 Conclusions......Page 368 References......Page 369 Introduction......Page 370 Related Work......Page 371 New User/Resource......Page 372 Semi-supervised Relational Methods......Page 373 Experiments......Page 375 Experiment Setting......Page 376 Results and Discussion......Page 377 Conclusions and Future Work......Page 378 References......Page 379 Introduction......Page 380 Normalized Discount Cumulative Gain (NDCG)......Page 381 Analysis of Listwise Loss Function......Page 382 Cost-Sensitive Listwise Loss Function......Page 383 Differences from Cost-Sensitive Ranking SVM......Page 384 Experiments......Page 385 Ranking Accuracy of ListMLE and Cost-Sensitive ListMLE......Page 386 Conclusion......Page 387 A Theoretical Justification of Lemma 1......Page 388 Introduction......Page 389 Question Expansion......Page 390 Expansion with Wikipedia......Page 391 Expansion with Yahoo! Answers......Page 392 Experimental Setup......Page 393 Performance Evaluation......Page 394 Conclusion and Future Work......Page 395 References......Page 396 Introduction......Page 397 Related Work......Page 398 The Proposed Method......Page 399 Data Sets......Page 400 Evaluation Settings......Page 401 Recall-by-Length Performance Curves......Page 402 References......Page 403 Introduction......Page 405 Related Work......Page 406 Divergence Feature for Vocabulary Filtering......Page 407 Three-Level Filtering for Term Weighting......Page 408 Archived Question Search with Vocabulary Filters......Page 409 Conclusions......Page 411 References......Page 412 Introduction......Page 413 Latent Dirichlet Allocation......Page 414 Distribution over Singular Values......Page 415 Topic Splitting......Page 416 Norms in Higher Dimension......Page 417 Divergence Measure......Page 418 Image Data......Page 419 Text Data......Page 420 Discussions and Conclusion......Page 422 References......Page 423 Introduction......Page 425 Latent Topic Model......Page 426 Max-margin Classification and Regression......Page 427 Framework of Supervising Latent Topic Model......Page 428 Variational Upper Bounds for Expected Max-margin Loss Functions......Page 429 Optimization Procedure for Classification......Page 430 Implementation......Page 431 Optimization Procedure for Text Regression......Page 432 State of Art Approaches......Page 433 Text Classification......Page 434 Text Regression......Page 435 References......Page 436 Introduction......Page 437 Motivation......Page 438 Information Extraction from the Web......Page 439 Active Information Acquisition......Page 440 Overview of the Architecture......Page 441 Query Engine......Page 442 Information Extraction......Page 443 Uncertainty Propagation in Citation Graph......Page 444 Confidence Evaluation System......Page 445 Results and Discussion......Page 446 Conclusion and Future Work......Page 448 References......Page 449 Introduction......Page 450 Related Work......Page 452 Reinforcement Learning Framework......Page 453 Reward Calculation......Page 454 Q-Value Approximation and Surfacing Algorithm......Page 456 Experiments......Page 458 Performance Comparison with Baseline Method......Page 459 Conclusion and Future Work......Page 460 References......Page 461 Introduction......Page 462 Related Work......Page 463 Topic Decomposition Based on NMF......Page 464 Topic, Sub-topics and Incidents Summarization......Page 465 Experimental Studies......Page 466 Summarization Evaluations......Page 467 Topic Decomposition......Page 468 References......Page 469 Introduction......Page 471 Related Works......Page 472 Problem Definition......Page 473 Uncertain Perceptron......Page 474 Algorithm Analysis......Page 476 Network Training......Page 477 Improve on Activate Function......Page 478 Experiment on Accuracy......Page 479 References......Page 481 Introduction......Page 483 Related Work......Page 484 Theoretical Background......Page 485 SkyDist by Monte-Carlo Sampling......Page 486 A Sweep-Plane Approach for the High-Dimensional Case......Page 487 Efficiency......Page 489 Clustering Skylines of Real World Data......Page 490 References......Page 492 Introduction......Page 493 Our Approach......Page 494 Query Processing......Page 495 Performance Study......Page 499 Conclusions......Page 500 References......Page 501 Introduction......Page 502 Related Work......Page 503 Sampling the Uncertain Dataset......Page 504 Statistical Bounds on the Quality of the Approximation......Page 505 Experiments......Page 506 References......Page 509 Introduction......Page 510 NS-PDT for Uncertain Data......Page 511 Static Classifier Ensemble......Page 512 Dynamic Classifier Ensemble......Page 513 Experiments......Page 514 Moving Hyperplane Concept with Gradual Concept Drift......Page 515 Time Analysis......Page 516 References......Page 517 Author Index......Page 518