Shinichiro Yoshii

Finding Unknown Interests Utilizing the Wisdom of Crowds in a Social Bookmark Service

In this paper, we develop an information recommendation system based on information edited by various users. This system targets a social bookmark service. Our method for finding appropriate information is to gather each users bookmark data on the Web and develop a system that analyzes the users association. Assuming that users that have similar interests also have information associated to those interests, we describe user co-citation relationships as a network. We extract the sub-graph as the community in which each user centers from this network. And we develop a Web service designed to automatically recommend new information regarding these interests. That service has a list of users who share similar preferences with the specified user and a system to present recommended information to the user

Spectrum Management of DSL Systems in Japan

The Japanese broadband market is maturing due to expansion in the needs for richer IT lifestyles with higher bit rate access. DSL technologies have played an important role in such growth because of their lower cost and ease of deployment. In order to expand the coverage area and compete with faster services such as FTTH, the DSL family of technologies is diversifying, and there are various services on the market currently in Japan. It is very important to manage those different kinds of systems in order to minimize interference due to crosstalk. This paper describes the Japanese spectrum management standard put into effect this year.

Self-organized criticality on growing scale-free networks

This paper explores a universal property in the behavior of growing scale-free networks. The characteristic of scale-free networks is that the degree distribution follows the power-law. This structure has been found in various kinds of self-organized networks. Most investigations conducted so far have demonstrated that network topologies are scale-free at a specific point in time. On the other hand, we focus attention on universality in the growing process of networks. In our proposed model, each node has its own fitness to designate the tendency allowing the node to acquire new links. From the simulation results, spread of the network follows the power-law, and power spectrum of the growing process shows 1/f noise, not to mention that the network has scale-free structure. It is found that those properties are in common with self-organized criticality. In conclusion, self-organizational growing networks follow the power-law not only in the sense of scale-free characteristic but also in the spatial and temporal sense.

Two types of adaptive multi-agent system

A methodology for multi-agent system communication is proposed. For knowledge acquisition of multi-agents system by learning, it is very important to pick up partial agent groups that should own rewards among themselves. In the proposed learning scheme, agents acquire a communication protocol. By this communication, the agents can make a particular organism that an ordinary reactive agent without communication can not generate.