Invited Papers -- Challenge for Info-plosion -- Machine Learning in Ecosystem Informatics -- Simple Algorithmic Principles of Discovery, Subjective Beauty, Selective Attention, Curiosity & Creativity -- A Theory of Similarity Functions for Learning and Clustering -- A Hilbert Space Embedding for Distributions -- Long Papers -- Time and Space Efficient Discovery of Maximal Geometric Graphs -- Iterative Reordering of Rules for Building Ensembles Without Relearning -- On Approximating Minimum Infrequent and Maximum Frequent Sets -- A Partially Dynamic Clustering Algorithm for Data Insertion and Removal -- Positivism Against Constructivism: A Network Game to Learn Epistemology -- Learning Locally Weighted C4.4 for Class Probability Estimation -- User Preference Modeling from Positive Contents for Personalized Recommendation -- Reducing Trials by Thinning-Out in Skill Discovery -- A Theoretical Study on Variable Ordering of Zero-Suppressed BDDs for Representing Frequent Itemsets -- Fast NML Computation for Naive Bayes Models -- Unsupervised Spam Detection Based on String Alienness Measures -- A Consequence Finding Approach for Full Clausal Abduction -- Literature-Based Discovery by an Enhanced Information Retrieval Model -- Discovering Mentorship Information from Author Collaboration Networks -- Active Contours as Knowledge Discovery Methods -- An Efficient Polynomial Delay Algorithm for Pseudo Frequent Itemset Mining -- Discovering Implicit Feedbacks from Search Engine Log Files -- Regular Papers -- Pharmacophore Knowledge Refinement Method in the Chemical Structure Space -- An Attempt to Rebuild C. Bernard's Scientific Steps -- Semantic Annotation of Data Tables Using a Domain Ontology -- Model Selection and Estimation Via Subjective User Preferences -- Detecting Concept Drift Using StatisticalTesting -- Towards Future Technology Projection: A Method for Extracting Capability Phrases from Documents -- Efficient Incremental Mining of Top-K Frequent Closed Itemsets -- An Intentional Kernel Function for RNA Classification -- Mining Subtrees with Frequent Occurrence of Similar Subtrees -- Semantic Based Real-Time Clustering for PubMed Literatures.

This volume contains the papers presented at DS-2007:The Tenth International Conference on Discovery Science held in Sendai, Japan, October 1-4, 2007. The main objective of the Discovery Science (DS) conference series is to p- vide an open forum for intensive discussions and the exchange of new ideas and information among researchers working in the area of automating scienti?c d- covery or working on tools for supporting the human process of discovery in science. It has been a successful arrangement in the past to co-locate the DS conference with the International Conference on Algorithmic Learning Theory (ALT). ThiscombinationofALTandDSallowsforacomprehensivetreatmentof the whole range, from theoretical investigations to practical applications. C- tinuing this tradition, DS 2007 was co-located with the 18th ALT conference (ALT 2007). The proceedings of ALT 2007 were published as a twin volume 4754 of the LNCS series. The International Steering Committee of the Discovery Science conference series provided important advice on a number of issues during the planning of Discovery Science 2007. The members of the Steering Committee are Einoshin Suzuki (Kyushu University, Chair), Achim G.