Journal of the Japan Society of Erosion Control Engineering, Vol.70, No.1, 2017

Occurrence and non]occurrence of the extraction conditions
and the occurrence of deep]seated landslide important factor using rough set

Soichi KAIHARA, Atsuhiko KINOSHITA, Naoki IMAMORI, Hiroaki TAKEMOTO and Noriko TADAKUMA

Numerous deep]seated landslides have occurred in southern Wakayama and Nara prefectures, with over 1,000mm of rainfall caused by typhoon no. 12 in 2011, which resulted in a great deal of damage. Understanding the occurrence and non]occurrence of deep]seated landslide conditions, it may be helpful in risk assessment and soft measures in the region, laborsaving field survey. As important factors that influence the occurrence or non]occurrence of deep]seated landslide, it may be slope direction, number of hits micro]topography, probable excess total rainfall, difference in height in the watershed, Difference of summit level and the lowest elevation, Distance from the fault. So, we studies to understand the relationship between incidence and non]incidence factors to be considered in the present study is closely related to deep]seated landslide. In addition, we extract the important factor in considering the occurrence or non]occurrence of deep]seated landslide using rough set. We have investigated the non]occurrence rules of deep]seated landslide occurrence rule. The results of the analysis by rough set, the number of points that can be described in a single rule, but there are few on the occurrence of deep]seated landslide, is that the description of about1/3of the total number of basin generation is possible and fit some rules All right. Further, as the non]generation rules, the rules to be little non]occurrence condition is extracted. Was applied for the first time to the occurrence of deep]seated landslide rough set was used to analyze the conditions of occurrence and non]occurrence of shallow landslides and debris flow. We were found to be useful in the understanding of the conditions and rules extraction of the occurrence of deep]seated landslide objective.

Key wordsFdeep]seated landslide, rough set, data mining, risk assessment


Japan Society of Erosion Control Engineering
Sabo Kaikan Bekkan 3F, 2-7-5 Hirakawa-cho, Chiyoda-ku, Tokyo
Phone +81-3-3222-0747 Fax +81-3-3230-6759
http://www.jsece.or.jp/
Mail jimu@jsece.or.jp