Vessel knowledge tag in Grid
In Grid, a vessel refers to a substantial watercraft, such as a ship or large boat. Vessels present a maritime compliance risk as they can be used as vehicles to circumvent sanctions.
#Vessel knowledge tag
Many risk codes within Grid can be related to vessel profiles. Events that involve vessels are prevalent in monitored lists and media events and are frequently designated with the tag #Vessel. Those often include vessels identified by an International Maritime Organization (IMO) number or any reference to a vessel, boat, ship, or other similar maritime vehicles.
For example:
“This entity appears in the National Institute of Japan External Trade Organization List on suspicion of illegal ship-to-ship trading with <country name>-related vessels and is subject to independent sanctions by United States. Identification Type: IMO Number Value: 8808264”
“<company name>, a Japanese fishing company, was convicted and sentenced to pay a fine of $1.5 million for two violations of the Act to Prevent Pollution from Ships and one count of obstruction of an agency proceeding. The charges stemmed from discharges of waste oil and oily bilge water from the <vessel name> No. ### (the vessel) into international waters and the attempt to cover up those discharges when the vessel was inspected by the U.S. Coast Guard in Apra Harbor, Guam.”
Knowledge tag quality control
Diligent quality control (QC) testing is integral to knowledge tags. Each new knowledge tag takes a staged approach to implementing quality control measures.
The stages of knowledge tag quality control are as follows:
Manual analyst teams are provided with instructions for knowledge tags, followed by a review period prior to tagging.
Tagging begins with phases of quality control measures, starting with 100% of all applied tags reviewed.
Tagging is incorporated into standard QC. If, at any point, QC results are less than 98%, the process is moved back a step to include additional QC to determine the cause of the decline in quality.
In the final stage, the standard quality control procedures are incorporated. This includes a focused monthly review based on trend analysis of unlikely risk code and knowledge tag combinations. The rate of QC is dependent on the volume of these occurrences.
Restricted and unrestricted knowledge tags
Knowledge tags are restricted to one Grid risk code for use cases within a subset of an existing risk code.
Knowledge tags are unrestricted to Grid risk codes for use cases that may intersect with multiple existing risk codes. Associative trend analysis can inform commonly expected risk code and knowledge tag combinations and is leveraged in the quality control process to provide an additional review for edge cases.