Large-Scale Knowledge Resources. Construction and Application

Atthestartofthe21stcentury,wearenowwellonthewaytowardsaknowled- intensive society, in which knowledge plays ever more important roles. Thus, research interest should inevitably shift from information to knowledge, with the problems of building, organizing, maintaining and utilizing knowledge - coming centralissues in a wide varietyof ?elds. The 21stCentury COE program Framework for Systematization and Application of Large-scale Knowledge - sources (COE-LKR) conducted by the Tokyo Institute of Technology is one of several early attempts worldwide to address these important issues. Inspired by this project, LKR2008 aimed at bringing together diverse contributions in cognitive science, computer science, education and linguistics to explore design, construction, extension, maintenance, validation and application of knowledge. Respondingtoourcallforpapers,wereceived38submissionfromavarietyof researchareas.EachpaperwasreviewedbythreeProgramCommitteemembers. Since we were aiming at an interdisciplinary conference covering a wide range of topics concerning large-scale knowledge resources (LKR), each paper was assigned a reviewer from a topic area outside the main thrust of the paper. This reviewer was asked to assess whether the authors described the moti- tion and importance of their work in a comprehensible manner even for readers in other research areas. Following a rigorous reviewing process, we accepted 14 regular papers and 12 poster papers.

Inhalt
Keynote Talk.- From Information to Intelligence: The Role of Relative Significance in Decision Making and Inference.- Mining Knowledge.- Comparing LDA with pLSI as a Dimensionality Reduction Method in Document Clustering.- Identification of MCMC Samples for Clustering.- TGC-Tree: An Online Algorithm Tracing Closed Itemset and Transaction Set Simultaneously.- Extracting Concepts from Religious Knowledge Resources and Constructing Classic Analysis Systems.- What Types of Translations Hide in Wikipedia?.- Initial Solution Set Improvement for a Genetic Algorithm in a Metadata Generation Support System for Landscape Photographs.- Identifying Semantic Relations in Japanese Compound Nouns for Patent Documents Analysis.- Extracting Prehistories of Software Refactorings from Version Archives.- Building Resources.- Initiatives, Tendencies and Driving Forces for a Lexical Web as Part of a Language Infrastructure.- Corpus Annotation/Management Tools for the Project: Balanced Corpus of Contemporary Written Japanese.- Capturing the Structures in Association Knowledge: Application of Network Analyses to Large-Scale Databases of Japanese Word Associations.- Construction of a Probabilistic Hierarchical Structure Based on a Japanese Corpus and a Japanese Thesaurus.- CHISE: Character Processing Based on Character Ontology.- Systematization of Knowledge about Performative Verbs: Capturing Speaker's Intention.- Toward Construction of a Corpus of English Learners' Utterances Annotated with Speaker Proficiency Profiles: Data Collection and Sample Annotation.- Filling the Gap between a Large-Scale Database and Multimodal Interactions.- Design and Prototype of a Large-Scale and Fully Sense-Tagged Corpus.- Image and Video.- Soccer Formation Classification Based on Fisher WeightMap and Gaussian Mixture Models.- Supervised Learning of Similarity Measures for Content-Based 3D Model Retrieval.- Automatic Score Scene Detection for Baseball Video.- Using Resources.- Large Scale e-Language Laboratory Based on Web 2.0.- Distant Collocations between Suppositional Adverbs and Clause-Final Modality Forms in Japanese Language Corpora.- Using Singular Value Decomposition to Compute Answer Similarity in a Language Independent Approach to Question Answering.- On the Representation of Perceptual Knowledge for Understanding Reference Expressions.- A Computational Model of Risk-Context-Dependent Inductive Reasoning Based on a Support Vector Machine.- Order Retrieval.- Stylistic Analysis of Japanese Prime Ministers' Diet Addresses.- Evaluation of Logical Thinking Ability through Contributions in a Learning Community.- Infrastructure.- Web Architecture and Naming for Knowledge Resources.- Towards Better Evaluation for Human Language Technology.- An Effective Scheduling Scheme for Information Searching with Computational Resources Scattered over a Large-Scale Network.