The Global Fishing Watch detection algorithm is a best effort to mathematically identify “apparent fishing activity” by using machine learning at a global scale. It is based on thousands of “training segments” that humans with expertise in fishing have manually classified. Just like humans, the algorithm will make some mistakes. Over time, this will continue to improve as we feed the algorithms more training data and correct the mistakes that are made. Further scrutiny is required before any legal action or proof can be made. Overall, the model can recreate the assessment of expert labelers with more than 90 percent accuracy.
Any and all references to “fishing” should be understood in the context of Global Fishing Watch’s fishing detection algorithm, which is a best effort to determine “apparent fishing effort” based on data from the Automatic Identification System (AIS) collected via satellites and terrestrial receivers. As AIS data varies in completeness, accuracy and quality, it is possible that some fishing effort is not identified and conversely, that some fishing effort identified is not fishing. For these reasons, Global Fishing Watch qualifies all designations of vessel fishing effort, including synonyms of the term “fishing effort,” such as “fishing” or “fishing activity,” as “apparent,” rather than certain. Any/all Global Fishing Watch information about “apparent fishing effort” should be considered an estimate and must be relied upon solely at your own risk. Global Fishing Watch is taking steps to make sure fishing effort designations are as accurate as possible.