One effort to prevent forest fires is to determine the appearance of hotspots as indicators of forest fires in a region. Sequential patterns of hotspot occurrences can be extracted from a hotspot dataset. Based on sequential patterns, we can know some regions where forest fires may potentially occur. In addition, we can know the time interval of hotspot occurrences in the region. Such information can be used to make decisions to prevent the forest fires. This work applies the sequential pattern mining algorithm namely PrefixSpan to find frequent sequences in the hotspot dataset in Riau from 2000 to 2014. We utilized the Sequential Pattern Mining Framework (SPMF) tool to generate sequences on hotspots data. Using the dataset of the year 2005 and the minimum support of 1% to 11%, we obtain 67 one-frequent sequences, 46 two-frequent sequences, and 1 three-frequent sequence. The sequential pattern with 2 items and minimum support of 2% shows that there were 178 hotspots sequentially occurred on Feb 10, 2005 then on Feb 12, 2005. The time interval of hotspot occurrences is of 3 days.
[1]
Sourav S. Bhowmick,et al.
Sequential Pattern Mining: A Survey
,
2003
.
[2]
Jian Pei,et al.
Sequential Pattern Mining by Pattern-Growth: Principles and Extensions
,
2005
.
[3]
Lailan Syaufina,et al.
Hotspot Distribution Analyses Based on Peat Characteristics Using Density-based Spatial Clustering
,
2015
.
[4]
Jianyong Wang,et al.
Mining sequential patterns by pattern-growth: the PrefixSpan approach
,
2004,
IEEE Transactions on Knowledge and Data Engineering.
[5]
Indry Dessy Nurpratami,et al.
Classification Rules for Hotspot Occurrences Using Spatial Entropy-based Decision Tree Algorithm
,
2015
.
[6]
S. M. García,et al.
2014:
,
2020,
A Party for Lazarus.