One massive problem with data collection on fishing, is that the world is absolutely littered with outright fake AIS data, where the vast majority are registered as Chinese vessels. Anyone can purchase a AIS transmitter, and spoof the data. Some areas are much more affected - especially out in the middle of the pacific ocean, and the Indian sea. Some areas there can have almost 100% spoofed AIS data. So a lot of work goes toward filtering these out.
When working on this kind of analysis, you have a bunch of data sources: AIS, VMS, LRIT, which are either land or satellite borne. Other than these, you have SAR, Optical, NRD (navigation radar detectors), and some other - but with these, you obviously need to have some classification and correlation system. AIS is by far the most common source, and it is also the one that is easiest to manipulate.
Vessels can simply turn off their AIS transponders, while out at sea. And it can easily be spoofed. But a lot of the garbage AIS data is really just that, random garbage. Just some random MMSI attached to a Chinese flag, and a completely random sailing pattern. These are relatively easy to filter out, but often times they share the MMSI with actual vessels - many which are indeed Chinese vessels.