Abstract

In recent years, the increased demand in the fishing industry has led to indiscriminate exploitation of maritime resources. To address this problem, Chile established boat fishing quotas that are inspected on-site but at random, verifying that their cargo corresponds to the tons of fishes indicated in their reports. However, this is inefficient and fraudulent fishermen have not been identified, increasing the necessity of a focused inspection. This paper proposes a flags system based on statistical data analysis, able to find potential fraudulent fishing boats. The flag system is built based on several hypothesis tests that evaluate significant differences in two aspects: fish frequency and cargo container utilization, where any significant difference is considered a new flag. The results show multiple boats with different behaviors for on-site versus documented reports, implying that the proposed flag system can point out fishing vessels with highly suspicious behavior.