Osamu Konishi

Mining of Topographic Feature from Heterogeneous Imagery and Its Application to Lunar Craters

In this study, a crater detection system for a large-scale image database is proposed. The original images are grouped according to spatial frequency patterns and both optimized parameter sets and noise reduction techniques used to identify candidate craters. False candidates are excluded using a self-organizing map (SOM) approach. The results show that despite the fact that a accurate classification is achievable using the proposed technique, future improvements in detection process of the system are needed.

Temporal Rule Discovery for Time-Series Satellite Images and Integration with RDB

Feature extraction and knowledge discovery from a large amount of image data such as remote sensing images have become highly required recent years. In this study, a framework for data mining from a set of time-series images including moving objects was presented. Time-series images are transformed into time-series cluster addresses by using clustering by two-stage SOM (Self-organizing map) and time-dependent association rules were extracted from it. Semantically indexed data and extracted rules are stored in the object-relational database, which allows high-level queries by entering SQL through the user interface. This method was applied to weather satellite cloud images taken by GMS-5 and its usefulness was evaluated.

Data Mining for Distributed Databases with Multiagents

We propose a technique for using multiagent technology in data mining intended for two or more text databases. In this paper, we discuss data mining method based on text (text mining), but our proposed method is not a method of specializing in text mining. First of all, we introduce some typical techniques as a technique of data mining to text database. Next, multiagent technology is described. We propose data mining technique using multiagent technology. The proposed technique is applied to document databases, and discuss its results.

Extension of Multiagent Data Mining for Distributed Databases

We proposed a technique for using multiagent technology in data mining intended for two or more text databases. In this paper, we discuss data mining method based on text (text mining), but our proposed method is not a method of specializing in text mining. First, we proposed data mining technique using multiagent technology. The proposed technique is applied to document databases, and discuss its results. Next, we extend the proposed technique with Stem algorithm, English morphological analysis, changed development language, adding the experiment data, and adding data mining algorithm.

Construction of school temperature measurement system with sensor network

We propose the sensor network system using the microcomputer board that can connect to the Internet. This proposed system can acquire information from the sensor of the microcomputer group arranged on the network, and can view collected information on Web browser. In this paper, it is shown to be able to construct easily the microcomputers sensor network which is combined microcomputer modules (Micro Cube) and the database server and the Web application server. The system that measured the room temperature in school campus was constructed, it has run for four months, and the effectiveness is verified.

Distributed web integration with multiagent data mining

We proposed a technique for using multiagent technology in data mining intended for two or more text databases. In this paper, we discuss data mining method based on text (text mining), but our proposed method is not a method of specializing in text mining. First, we proposed data mining technique using multiagent technology. The proposed technique is applied to document databases, and discuss its results. In this paper, proposed data mining using multiagent was applied to information integration system on Web, and the effectiveness was verified. In the proposed method, the part of the database access agent was changed to the Web access agent. Also, mining agent was changed to the information extraction agent from the HTML file.

Social bookmark data mining using extended graph kernel

In this study, we present a social bookmark data community and propose a method to extract the rules of community changes. The transaction data collected from a data stream is considered a graph representing change in structure and sequence data for each relevant time period, while analyzing changes in the sequence graph of the community. The algorithm proposed in this paper is the hierarchical clustering method combined with convolution graph kernels weighted in time. We use this algorithm on the entire community to analyze the relationship among graph sequences and show that it occasionally appears (disappearing in the middle of the sequences) and changes the extracted community rules. Experimental results obtained using synthetic data sets and real social bookmark data show that changes in the community captured the occasional occurrence of the proposed algorithm. The proposed method can detect and analyze social bookmark data including large-scale time-series graph.

Peer-to-peer sensor network system for a school temperature measurement system

In this paper, we propose the technique of sensor data mining by the peer-to-peer (P2P) network. The mechanism that it is possible to share on the P2P network is considered by receiving information from the sensor by the P2P application. A searching request for a sensor unit and mining the sensor data occurs on the P2P application. We applied the proposed technique to a school environment measurement system. In this system, sensor units are arranged on campus and a user can measure a room’s temperature and humidity. The temperature sensor and the humidity sensor are implemented in a microcomputer board that can connect to the Internet, and we define the microcomputer board as a sensor unit. We construct the P2P sensor network on which a PC accesses the sensor unit and the P2P application on its PC uploads on the P2P network. The P2P network can disclose sensor information after more advanced processing is given by thinking as a P2P application and not a sensor unit, but on the sensor unit and the PC.

Broiler-House Environment Monitoring System using Sensor Network and Mail Delivery System

In this paper, we propose the system that combines the sensor network with the mail delivery system to construct the system that observes an environmental change of the broiler-house. As a result of hearing of the producer, the environment system needs to able to be observed some broiler-houses, to inspect the summary data from the cellular phone, and to transmit the warning mail in a rapid temperature change. A basic part of the system is a sensor network by the sensor module that we developed. Only the sensor modules are put in the each broiler-house, and the network by wireless LAN communication is constructed, because the system needs to watch of two or more broiler-houses, and it is difficult to setup a large-scale system at the broiler-house. The always-connected high-speed Internet is preferable to accumulate, to process data, and to offer it to the user in a comprehensible form. But, it is difficult to build always-connected high-speed Internet at the chicken farm which is used by experiment. The server is set up in the remote place, and we propose the system that delivers data from the chicken farm with mail. The verification of the effectiveness of the proposed system and the problem are examined by actually setting up the system that proposes it in the broiler-house, and operating it.

Applied Immune Algorithm to Search Optimum Compositions of Solid-state Catalysts

An immune algorithm is applied to search optimal concentrations of solid-state catalysts. In order to evaluate the activity of catalysts, an artificial neural network trained by the observed data is adopted. This method makes us possible to obtain potential catalyst composition with high activity.