§16§The Third International Conference on Advanced Data Mining and Applications (ADMA) organized in Harbin, China continued the tradition already established by the first two ADMA conferences in Wuhan in 2005 and Xi'an in 2006. One major goal of ADMA is to create a respectable identity in the data mining research com- nity. This feat has been partially achieved in a very short time despite the young age of the conference, thanks to the rigorous review process insisted upon, the outstanding list of internationally renowned keynote speakers and the excellent program each year. The impact of a conference is measured by the citations the conference papers receive. Some have used this measure to rank conferences. For example, the independent source cs-conference-ranking.org ranks ADMA (0.65) higher than PAKDD (0.64) and PKDD (0.62) as of June 2007, which are well established conferences in data mining. While the ranking itself is questionable because the exact procedure is not disclosed, i §16§t is nevertheless an encouraging indicator of recognition for a very young conference such as ADMA. §04§ng Metropolis-Hastings Algorithm.- A Consensus Recommender for Web Users.- Constructing Classification Rules Based on SVR and Its Derivative Characteristics.- Hiding Sensitive Associative Classification Rule by Data Reduction.- AOG-ags Algorithms and Applications.- A Framework for Titled Document Categorization with Modified Multinomial Naivebayes Classifier.- Prediction of Protein Subcellular Locations by Combining K-Local Hyperplane Distance Nearest Neighbor.- A Similarity Retrieval Method in Brain Image Sequence Database.- A Criterion for Learning the Data-Dependent Kernel for Classification.- Topic Extraction with AGAPE.- Clustering Massive Text Data Streams by Semantic Smoothing Model.- GraSeq: A Novel Approximate Mining Approach of Sequential Patterns over Data Stream.- A Novel Greedy Bayesian Network Structure Learning Algorithm for Limited Data.- Optimum Neural Network Construction Via Linear Programming Minimum Sphere Set Covering.- How Investigative Data Mining Can Help §04§Intelligence Agencies to Discover Dependence of Nodes in Terrorist Networks.- Prediction of Enzyme Class by Using Reactive Motifs Generated from Binding and Catalytic Sites.- Bayesian Network Structure Ensemble Learning.- Fusion of Palmprint and Iris for Personal Authentication.- Enhanced Graph Based Genealogical Record Linkage.- A Fuzzy Comprehensive Clustering Method.- Short Papers.- CACS: A Novel Classification Algorithm Based on Concept Similarity.- Data Mining in Tourism Demand Analysis: A Retrospective Analysis.- Chinese Patent Mining Based on Sememe Statistics and Key-Phrase Extraction.- Classification of Business Travelers Using SVMs Combined with Kernel Principal Component Analysis.- Research on the Traffic Matrix Based on Sampling Model.- A Causal Analysis for the Expenditure Data of Business Travelers.- A Visual and Interactive Data Exploration Method for Large Data Sets and Clustering.- Explorative Data Mining on Stock Data - Experimental Results and Findings.- Graph S §04§tructural Mining in Terrorist Networks.- Characterizing Pseudobase and Predicting RNA Secondary Structure with Simple H-Type Pseudoknots Based on Dynamic Programming.- Locally Discriminant Projection with Kernels for Feature Extraction.- A GA-Based Feature Subset Selection and Parameter Optimization of Support Vector Machine for Content - Based Image Retrieval.- E-Stream: Evolution-Based Technique for Stream Clustering.- H-BayesClust: A New Hierarchical Clustering Based on Bayesian Networks.- An Improved AdaBoost Algorithm Based on Adaptive Weight Adjusting.
