The traditional approach for determining potential fishing zones (PFZs) relies on oceanographic factors (biological, physical, and chemical) and fishermen's expertise. This approach has disadvantages particularly when it comes to the analysis of combining these factors to find an exact PFZ spatially and temporally. In this study, we proposed a framework for identifying PFZs based on a data-mining approach in the Eastern Indian Ocean. We utilized a spatio-temporal clustering method to identify clusters of zones with data on the largest number of fish catch, which were then integrated with the sea surface temperature (SST) and the sea surface chlorophyll a (SSC) data derived from Moderate Resolution Imaging Spectroradiometer (MODIS) satellite imagery. The results of this data integration method were used as training data in the classification process, which was then used to determine PFZs. During the classification process, we utilized the k-nearest neighbor (KNN) classification method. The result gave an average accuracy of 87.11%, which showed that the proposed framework can be used effectively to determine PFZs. To validate the framework, we compared its performance against the heuristic rules taken from the knowledge-based expert system model on the SST and chlorophyll a data. The results showed that the proposed data-mining framework outperformed the heuristic rules from the knowledge-based expert system model.
|Number of pages||9|
|Journal||IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing|
|Publication status||Published - Aug 2016|
- Data integration
- oceanographic characteristics
- potential fishing zone (PFZ)
- spatio-temporal data mining