Dataset compiled from spreading hot spots, responsible for fire risk in many regions of Indonesian forests, are complex, primarily induced by the large size of the observed regions and high variation of hot spot distribution. The challenge in analyzing this type of dataset is to develop statistical techniques that facilitate the analysis, visualization, and interpretation of the results. Techniques, such as multivariate analysis and artificial neural networks, have been applied to resolve the high-dimensional space in such large datasets. Each method uses a different rationale for how the relationship between the input parameters will be preserved during analysis. This study presents the use of a principal component analysis (PCA) and a self-organizing map (SOM) to reduce the high dimensionality of the input variables and, subsequently to visualize the dataset into a two-dimensional (2-D) space. The results indicate that the first two principal components of the PCA provide a large percentage of cumulative variance to explain the data patterns. However, a comparison of the data projection, SOM is better suited than PCA in visualizing the fire-risk distribution in forests. The SOM color-coding and labeling also effectively visualized a classification system of fire risk via node clusters, in such a way that the fire risks level according to their hot spot locations in forest is easily interpreted.

View full abstract

In text mining of documents of a specific area, especially for generating a map of concepts or terms and a summary of concepts or terms, the quality of keywords strongly affects the results of analysis. A list of technical terms is available as keyword candidates. We can recognize terms in the corpus automatically using a scoring method based on statistics of compound nouns. However, because fractions of words or meaningless strings are also included in those term candidates, further selections are necessary. For such further selection, we consider a method to obtain overlapping terms between the two groups of terms that are extracted from two independent corpora of the same area. For the experimental selection of terms, three target areas are specified: livestock raising, fruit farming, and vegetable gardening. For each area, two groups of documents are collected. The term candidates are extracted from these corpora using a scoring method based on statistics of compound nouns. The terms overlapping the two groups are extracted. After this selection procedure, the proportion of unsuitable terms is lower. From an efficiency viewpoint, the selection procedure improves selection. In addition, the procedure provides the advantage that it is independent from subjective decisions related to manual selection.

View full abstract

Japonica rice production is central to Japanese agriculture. The main cause of production losses is the fungal disease rice blast, and its early detection and chemical control is indispensable for food safety and environmentally-friendly production. We have developed an automatic diagnostic technology for plant disease. This technique employs active sensing in virtual space using a 3-dimensional computer graphics simulation system (3DCG simulator), allowing visual detection of plant disease in virtual space. In this paper, a 3DCG simulator for rice blast was built. Next, a genetic programming based recognition algorithm was developed, and a rice blast detection technique using machine vision was examined in virtual space.

View full abstract

The grand design and the implementation of the farm planning support system FAPS-DB is presented. This system consists of four main components: the FAPS-DB integrated agro-technology system data base FSDB, a vegetable and fruit market information data base NAPASS, a farming index making system FMIGS and a farming technology system evaluation and planning system FAPS. FSDB and NAPASS both provide a SOAP based Web service interface that can be easily used from other Web applications. In this system, a farming index is interactively made on the Web, and can be displayed by graph and slit form. Moreover, the farming index data can be exported to two other systems: FSDBout and FAPS, which can use the data for various farming plan making matched to a user's purpose. The former system makes an introductory farming plan. The latter system makes an advanced farming plan that considers farming risk. As a result it has become possible to easily make a highly accurate farming plan to introduce a new crop or a new technology. User evaluation clarified that the following three improvements were necessary. Operation of the system needed to be simpler, price and quantity data needed to be able to be updated, and a link was required to the vegetable and fruit market data base. Adding these improvements to the system together with improvements to the screen design and ease of use have greatly improved system function. We have clarified that this system is likely to be used in agricultural extension centers and by farmers, research laboratories, and local governments.

View full abstract