Catálogo de publicaciones - libros

In this chapter, we proudly introduce eight awarded papers, selected from the papers presented in the 20th annual conference of Japan Society for Artificial Intelligence (JSAI2006).

The conference was held at Tower Hall Funabori located in the east part of Tokyo from June 7 until June 9, 2006. 276 papers were presented in about 60 sessions and over 500 people participated in the conference. Sessions vary from the fundamental issues to the state-of-the-art applications. Session for regular papers are as follows; Logic and learning, Reinforcement learning and agent learning, Agent theory, Auction/Game/Economics, Agent learning, Agent planning, Agent simulation and interaction, Genetic algorithm, Image processing, Information extraction and classification, Clustering/self-organization, Classifi- cation learning, Text mining, Graph mining, Mining algorithm, Web mining, Pre- and post-processing for data mining, Practices of data mining, Cognitive modeling, Language processing and dialogue, Robot/sensor network, Web information system, Semantic Web, Knowledge modeling and knowledge sharing, Support of knowledge management, knowledge modeling/ontology,Web service, Human interface and communication support, Education support, Learning support environment, and Musical and auditory information processing.

We treat the problem of subsequence time-series clustering (STSC) from a group-theoretical perspective. First, we show that the sliding window technique introduces a mathematical artifact to the problem, which we call the pseudo-translational symmetry. Second, we show that the resulting cluster centers are necessarily governed by irreducible representations of the translational group. As a result, the cluster centers necessarily forms sinusoids, almost irrespective of the input time-series data. To the best of the author’s knowledge, this is the first work which demonstrates the interesting connection between STSC and group theory.

This article describes a model for sustainable contents management, its visualization algorithms, and the implemented system, called sustainable knowledge globe (SKG). The focal point of our study is visualization using contour maps. The graphical representation of tree-structured contents increases in complexity with the number of contents. The contour map representations can briefly depict the arrangement and structure of contents in an archive. Three contour map representations are proposed assuming the importance of the arrangement design. Nesting, dendroidal, and island-like contours are amplified from the viewpoint of preservation of the structures and arrangements, in addition to supporting shape and texture design. The comparison and applications of the three algorithms are discussed.

Discussion mining is a preliminary study on gathering knowledge based on the content of face-to-face discussion meetings. To extract knowledge from discussion content, we have to analyze not only the surface arguments, but also semantic information such as a statement’s intention and the discussion flow during meetings. We require a discussion ontology for this information. This discussion ontology forms the basis of our discussion methodology and requires semantic relations between elements in meetings. We must clarify these semantic relations to build the discussion ontology. We therefore generated discussion content and analyzed meeting metadata to build the ontology.

We propose a method for predicting types of protein-protein interactions using a multiple-instance learning (MIL) model. Given an interaction type to be predicted, the MIL model was trained using interaction data collected from biological pathways, where positive bags were constructed from interactions between protein complexes of that type, and negative bags from those of other types. In an experiment using the KEGG pathways and the Gene Ontology, the method successfully predicted an interaction type (phosphorylation) to an accuracy rate of 86.1%.

In this paper, we describe and discuss problems in terms of the music formalization framework that we are developing. Currently, we aim at building a contents design framework for ordinary people in a formal approach. Based on the knowledge representation technology, an art piece and contents are represented by a feature structure and the high-level operations upon the feature structure are available. When we apply the framework to implementing a practical musical system, we face some problems related to algebra, knowledge representation, and/or implementation. These are the issues concerned with: a relative pseudo-complement introduced for increasing descriptive power, an abstraction level of the distance between musical fragments, and a formalization of musical operations in addition to that of musical relations.

The topic of getting information from sensor networks is becoming more important. Middleware will be required to manage various sensor data in future. We propose the Semantic Sensor Network (SS), which applies an ontology to sensor data and enables such data to be treated more easily. However, users have great difficulty to understand the information on the SS because its semantic data are complicatedly related with each other and the data and their relations evolve in time. The requirements for visualizing such information are; 1) the SS must give users the information appropriately, and 2) the SS must answer the users’ queries. To meet these requirements, we propose Viewlon, software for visualizing information on the SS. To give the information appropriately, information display levels and a timeline view are introduced. In addition the functionality of processing queries provides the required information to users interactively. We experimented with Viewlon in a simulation environment, and the results show that Viewlon has the ability to present information to users and to answer their queries correctly.

Design and acquisition method of environmental model are important issues for robots which act in the real world such as a daily life environment. Robots have to follow up changes of environment and re-construct the model because of the conditions in the real world environment would be changed constantly. Also the model for unobserved part would be needed and should be inferred when users instruct the robot using symbolic expressions. In this paper, we have focused on stochastic representation of environmental memory to realize smooth communication between humans and robots, and realtime memory management with ambiguities in the real world. We also show that the representation is effective to construct a cooperative task achievement system on intelligent robots.

Web-based knowledge sharing communities, which are supported by countless voluntary Internet users, are in widespread use in our lives. In one of these communities, Yahoo! Chiebukuro, one of the most famous web-based knowledge sharing communities in Japan, interpersonal communication among members is evoked as well as information exchange and accumulation. In this research, we conducted a questionnaire survey of Yahoo! Chiebukuro members. Based on 7989 survey samples, we intended to describe some features of community members, patterns of their participation in the community, and the reasons they joined the community. Results suggest that many members join the community and post their own information with the altruistic motive of helping others.

This was the third Logic and Engineering of Natural Language Semantics workshop. This year it was held at Funabori Tower Hall in Tokyo, Japan on June 5th and 6th of 2006. There were three invited talks, by Makoto Kanazawa, Chung-Min Lee, and Christopher Potts. In addition, the program committee (see Acknowledgements) selected fifteen abstracts from those submitted, for a total of eighteen talks.

This year’s theme was formal pragmatics. In recent years there have been a number of exciting developments in this area. Researchers have applied gametheoretical and utility-theoretic techniques to problems such as Gricean communication and relevance, implicature, and blocking; for instance, new work in multidimensional logic has given insight into conventional implicature; new formal techniques have been applied to discourse structure and coherence. The goal of this year’s workshop was to bring researchers in these areas together for discussion. This was an eminent success.