Method, device and readable storage medium for determining sensitive ship

Abstract

This application proposes a method, apparatus, and readable storage medium for determining sensitive vessels. The method includes: acquiring vessel name data for multiple vessels to be detected; extracting name-sensitive features and behavior-sensitive features for each vessel to be detected, and constructing a sensitive feature trie; acquiring a first sensitive vessel dataset based on feature retrieval conditions and the sensitive feature trie, and determining vessel spatial parameters; determining a second sensitive vessel dataset based on the vessel spatial parameters and a spatial range threshold; determining name-sensitive features and / or behavior-sensitive features based on the second sensitive vessel dataset; determining a vessel sensitivity intensity value based on the spatial clustering and feature matching degree of the second sensitive vessel dataset; determining a target sensitive vessel based on the vessel sensitivity intensity value and a sensitivity intensity level threshold, and outputting the sensitivity intensity level data of the target sensitive vessel. This application combines the quantitative determination of feature matching degree and spatial clustering degree to achieve multi-dimensional and accurate identification of sensitive vessels.

Description

Technical Field

[0001] This invention relates to the field of shipping and maritime monitoring technology, and more specifically, to a method, apparatus, and readable storage medium for identifying sensitive vessels. Background Technology

[0002] In the field of maritime data monitoring and network information content security, the identification of sensitive vessels based on the single dimension of vessel name cannot integrate the sensitive characteristics of vessel behavior, and there is a lack of quantitative classification of sensitive vessels, resulting in insufficient accuracy and targeting of supervision. Summary of the Invention

[0003] The present invention aims to solve at least one of the technical problems existing in the prior art or related art.

[0004] Therefore, a first aspect of the present invention provides a method for identifying sensitive ships.

[0005] A second aspect of the invention provides a device for identifying sensitive ships.

[0006] A third aspect of the invention provides another device for identifying sensitive ships.

[0007] The fourth aspect of this application proposes a readable storage medium.

[0008] In view of the above, a first aspect of the present invention provides a method for determining sensitive vessels, comprising: in response to a retrieval instruction for a sensitive vessel, acquiring vessel name data of multiple vessels to be detected; extracting name-sensitive features and behavior-sensitive features of each vessel to be detected based on the vessel name data; constructing a sensitive feature dictionary tree based on the name-sensitive features and behavior-sensitive features; performing sensitive feature matching between the feature retrieval conditions carried in the retrieval instruction and the sensitive feature dictionary tree to obtain a first sensitive vessel dataset; extracting vessel location information from the first sensitive vessel dataset to determine vessel spatial parameters; determining a second sensitive vessel dataset based on the vessel spatial parameters and a preset spatial range threshold; determining the name-sensitive features and / or behavior-sensitive features of the sensitive vessel to be determined based on the second sensitive vessel dataset; determining the feature matching degree based on the feature retrieval conditions, name-sensitive features, and / or behavior-sensitive features; determining a vessel sensitivity intensity value based on the spatial clustering degree and feature matching degree of the second sensitive vessel dataset, wherein the spatial clustering degree is used to characterize the degree of clustering of the sensitive vessels to be determined; determining a target sensitive vessel based on the vessel sensitivity intensity value and a preset sensitivity intensity level threshold, and outputting the sensitivity intensity level data of the target sensitive vessel.

[0009] This application constructs a dynamically updated sensitive feature dictionary by integrating the two dimensions of ship name and behavior. Combined with the quantitative determination of feature matching degree and spatial clustering degree, it realizes multi-dimensional accurate identification, hierarchical determination and automated screening of sensitive ships, which greatly improves the efficiency, accuracy and targeting of maritime supervision and effectively responds to evasive sensitive violations.

[0010] In some technical solutions of this application, the name-sensitive features include: sensitive words in a preset sensitive word library and spelling variations of sensitive words, spelling variations including at least one of character substitution, homophonic rewriting, separator insertion, and abbreviation with missing characters, and character substitution including: substitution of numbers and letters, and substitution of special symbols and characters; the behavior-sensitive features include at least one of the following: abnormal vessel loitering features, abnormal signal features, intrusion into sensitive areas features, territorial sea violation features, illegal fishing features, and features of being included in the sanctions list.

[0011] In some technical solutions of this application, constructing a sensitive feature dictionary based on name-sensitive features and behavior-sensitive features includes: standardizing or labeling the name-sensitive features and behavior-sensitive features to obtain a standardized sensitive feature set; and constructing a sensitive feature dictionary based on the standardized sensitive feature set.

[0012] In some technical solutions of this application, after constructing a sensitive feature dictionary based on name-sensitive features and behavior-sensitive features, the method includes: configuring a dynamic update triggering mechanism for the sensitive feature dictionary; and performing a dynamic update operation on the sensitive feature dictionary based on the dynamic update triggering mechanism to obtain the updated sensitive feature dictionary.

[0013] In some technical solutions of this application, the sensitive feature dictionary is configured with a dynamic update trigger mechanism. The dynamic update trigger mechanism includes adding new sensitive feature types, adjusting regulatory rules, or updating node-mounted data that has not been updated within the specified period. The update operations include adding nodes, adjusting weights, or cleaning up invalid nodes.

[0014] In some technical solutions of this application, the feature retrieval conditions include at least one of the following: a single name-sensitive feature, a single behavior-sensitive feature, a logical AND / OR of multiple name-sensitive features, a logical AND / OR of multiple behavior-sensitive features, and a combination matching of name-sensitive features and behavior-sensitive features.

[0015] In some technical solutions of this application, the ship space parameters include at least one of the following: actual distance between ships, spatial coverage of ship group, and ship density of ship group. The preset spatial range threshold is a graded threshold set by maritime supervision according to the regulatory needs of different sea areas.

[0016] A second aspect of this application provides an apparatus for determining sensitive vessels, comprising: a first acquisition module, a first extraction module, a first construction module, a second acquisition module, a first determination module, a second determination module, a third determination module, a fourth determination module, a fifth determination module, and a sixth determination module. The first acquisition module is used to acquire vessel name data of multiple vessels to be detected in response to a retrieval instruction for sensitive vessels. The first extraction module is used to extract name-sensitive features and behavior-sensitive features for each vessel to be detected based on the vessel name data. The first construction module is used to construct a sensitive feature dictionary tree based on the name-sensitive features and behavior-sensitive features. The second acquisition module is used to perform sensitive feature matching with the sensitive feature dictionary tree based on the feature retrieval conditions carried in the retrieval instruction to acquire a first sensitive vessel dataset. The first determination module is used to extract the first sensitive... The system employs a series of modules: a first module determines ship spatial parameters from ship location information in a ship dataset; a second module determines a second sensitive ship dataset based on the ship spatial parameters and a preset spatial range threshold; a third module determines the name-sensitive features and / or behavioral-sensitive features of the ships to be judged based on the second sensitive ship dataset; a fourth module determines the feature matching degree based on feature retrieval conditions, name-sensitive features, and / or behavioral-sensitive features; a fifth module determines the ship sensitivity intensity value based on the spatial clustering degree and feature matching degree of the second sensitive ship dataset, where spatial clustering degree characterizes the degree of clustering of the ships to be judged; and a sixth module determines the target sensitive ship based on the ship sensitivity intensity value and a preset sensitivity intensity level threshold, and outputs the sensitivity intensity level data of the target sensitive ship.

[0017] A third aspect of the present invention provides an apparatus for determining a sensitive vessel, comprising: a processor and a memory, wherein the memory stores a program or instructions, and the processor, when executing the program or instructions in the memory, implements the steps of the method for determining a sensitive vessel as described in any of the above-described technical solutions. Therefore, the apparatus for determining a sensitive vessel possesses all the beneficial effects of the method for determining a sensitive vessel as described in any of the above-described technical solutions.

[0018] A fourth aspect of the present invention provides a readable storage medium storing a program or instructions, which, when executed by a processor, implement the steps of the method for determining a sensitive vessel as described in any of the above-described technical solutions. Therefore, the readable storage medium possesses all the beneficial effects of the method for determining a sensitive vessel as described in any of the above-described technical solutions.

[0019] Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0020] The above and / or additional aspects and advantages of the present invention will become apparent and readily understood from the description of the embodiments taken in conjunction with the following drawings, in which:

[0021] Figure 1 This is a flowchart illustrating a method for determining sensitive vessels according to an embodiment of the present invention.

[0022] Figure 2 This is an overall architecture diagram of a sensitive ship detection system corresponding to a sensitive name with spatial characteristics and a ship detection method according to an embodiment of the present invention;

[0023] Figure 3 This is a schematic diagram of ship navigation status and sensitive target monitoring according to an embodiment of the present invention;

[0024] Figure 4 One of the schematic block diagrams of a sensitive vessel determination device according to an embodiment of the present invention;

[0025] Figure 5 This is a second schematic block diagram of a sensitive vessel identification device according to an embodiment of the present invention. Detailed Implementation

[0026] To better understand the above-mentioned objectives, features, and advantages of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be noted that, unless otherwise specified, the embodiments and features described in these embodiments can be combined with each other.

[0027] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and therefore the scope of protection of the invention is not limited to the specific embodiments disclosed below.

[0028] The following reference Figures 1 to 5 Methods, apparatus, and readable storage media for identifying sensitive ships according to some embodiments of the present invention are described.

[0029] like Figure 1 As shown, an embodiment of this application provides a method for determining sensitive vessels, the steps of which include:

[0030] Step 102: In response to the search command for sensitive vessels, obtain the vessel name data of multiple vessels to be detected;

[0031] Step 104: Based on the ship name data, extract the name-sensitive features and behavior-sensitive features for each ship to be detected;

[0032] Step 106: Construct a sensitive feature trie based on name-sensitive features and behavior-sensitive features;

[0033] Step 108: Match sensitive features with the sensitive feature trie based on the feature retrieval conditions carried by the retrieval instruction to obtain the first sensitive ship dataset;

[0034] Step 110: Extract ship position information from the first sensitive ship dataset and determine ship spatial parameters;

[0035] Step 112: Determine the second sensitive ship dataset based on the ship's spatial parameters and the preset spatial range threshold;

[0036] Step 114: Based on the second sensitive vessel dataset, determine the name-sensitive features and / or behavioral-sensitive features of the vessels to be judged as sensitive.

[0037] Step 116: Determine the feature matching degree based on the feature retrieval conditions, name-sensitive features, and / or behavior-sensitive features;

[0038] Step 118: Determine the sensitivity intensity value of the ships based on the spatial clustering degree and feature matching degree of the second sensitive ship dataset, wherein the spatial clustering degree is used to characterize the degree of clustering of the sensitive ships to be determined;

[0039] Step 120: Based on the ship's sensitivity intensity value and the preset sensitivity intensity level threshold, determine the target sensitive ship and output the sensitivity intensity level data of the target sensitive ship.

[0040] The method for identifying sensitive vessels proposed in this application utilizes a vessel monitoring system. Upon responding to a search command for sensitive vessels initiated by maritime regulatory personnel, it acquires vessel name data for multiple vessels to be monitored, ensuring comprehensive screening coverage. Next, based on the acquired vessel name data, it extracts name-sensitive features and behavioral-sensitive features for each vessel to be monitored. Name-sensitive features are derived from matching the vessel name with a pre-defined sensitive word database, covering sensitive words and spelling variations such as character substitutions and homophonic alterations. Behavioral-sensitive features are determined by analyzing dynamic data from the Automatic Identification System (AIS) and maritime regulatory business data, covering various sensitive behaviors such as abnormal loitering and signal anomalies. This achieves multi-dimensional coverage of sensitive vessel information, avoiding the limitations of focusing solely on the name dimension. Then, based on the extracted name-sensitive features and behavior-sensitive features, a sensitive feature trie is constructed. Since the trie uses a multi-branch tree structure, with standardized single character or behavior feature label codes as nodes, the path from the root node to the terminal node constitutes a complete sensitive feature string. Sensitive feature strings with a common prefix share corresponding nodes. This not only continues the technical advantages of trie matching—high efficiency, memory saving, and ease of multi-pattern matching—but also achieves the fusion storage and retrieval of dual-dimensional sensitive features, solving the problem of low real-time matching efficiency for massive ship data. Subsequently, sensitive feature matching is performed between the feature retrieval conditions carried by the retrieval command and the sensitive feature trie, filtering to obtain a first sensitive ship dataset possessing the target sensitive features. This completes the initial filtering from a massive number of ships to be detected to a sensitive ship candidate set, significantly reducing the data scale for subsequent processing and improving overall screening efficiency. Afterward, ship location information is extracted from the first sensitive ship dataset to determine ship spatial parameters. These parameters are compared with preset spatial range thresholds to obtain a second sensitive ship dataset that possesses both the target sensitive features and meets the spatial range requirements. Through this secondary spatial filtering, sensitive ships clustered in local sea areas are accurately identified. Next, based on the second sensitive vessel dataset, the name-sensitive features and / or behavioral-sensitive features of the vessels to be judged are determined. The feature matching degree is calculated by combining feature retrieval conditions. Simultaneously, the spatial clustering degree is determined based on the normalized values ​​of the vessel's spatial parameters. Feature matching measures the degree of conformity of the vessel's sensitive features, and spatial clustering measures the degree of spatial concentration of the vessels, providing a two-dimensional indicator for the quantitative determination of sensitivity intensity. Then, based on the spatial clustering degree and feature matching degree of the second sensitive vessel dataset, the vessel sensitivity intensity value is calculated using a pre-defined quantitative formula. This quantitatively integrates the degree of conformity of the vessel's sensitive features with its spatial distribution status, achieving an upgrade from qualitative identification to quantitative judgment and solving the problem of how to measure sensitivity intensity.

[0041] Finally, based on the ship's sensitivity intensity value and the preset sensitivity intensity level threshold, the target sensitive ships are identified, and the sensitivity intensity level data of the target sensitive ships and related auxiliary information are output. This provides maritime regulatory personnel with accurate regulatory targets and a basis for graded handling, improving the pertinence and efficiency of supervision. It forms a complete logical closed loop of "data acquisition, dual feature extraction, trie construction, feature matching and filtering, spatial filtering, intensity quantification, and result output", which progressively realizes the accurate identification and graded determination of sensitive ships. The entire process is automated, which greatly improves the intelligence level of maritime supervision.

[0042] The sensitive vessel search command is an operational command initiated by maritime regulatory personnel based on monitoring and sensitive information supervision needs. It can carry characteristic search conditions and serves as the starting signal for the system to conduct sensitive vessel identification and screening, supporting screening of vessels in target sea areas or across the entire area. The vessel name data to be detected is the data related to the names of vessels requiring sensitive screening within the system. It originates from publicly available vessel dynamic information collected by AIS and is the basic identity data for extracting sensitive vessel features. Name sensitive features are words and spelling variations (character replacements, homophonic rewritings, etc.) extracted from the vessel name from a pre-set sensitive word library. These are the core features reflecting the vessel name's violations and sensitive attributes. Behavioral sensitive features are violations that conform to pre-set judgment rules during vessel navigation and operations, including abnormal loitering, abnormal signals, intrusion into sensitive areas, territorial waters violations, illegal fishing, and inclusion on a sanctions list. These are derived from the analysis of AIS dynamic data and maritime regulatory business data. The sensitive feature trie is a multi-branch tree retrieval structure built upon standardized name / behavioral sensitive features. Nodes are character or behavioral feature labels, paths are complete sensitive feature strings, and common prefixes share nodes. It boasts advantages such as high matching efficiency, low memory consumption, and support for multi-pattern matching, making it the core structure for multi-dimensional sensitive feature matching. Feature retrieval conditions are regulatory screening conditions carried in the retrieval command, which can flexibly combine single / multiple name / behavioral sensitive features with AND / OR relationships to adapt to different regulatory scenarios. The first sensitive vessel dataset is a set of vessels with target sensitive features selected after matching the feature retrieval conditions with the sensitive feature trie. It includes complete AIS data such as vessel name, location, and MMSI (Maritime Mobile Service Identity), serving as a preliminary candidate set of sensitive vessels in the feature dimension. Vessel location information is real-time geographic information of vessels collected by AIS, including latitude and longitude, and is the core data for calculating vessel spatial parameters and conducting spatial dimension screening. Vessel spatial parameters are spatial distribution indicators calculated based on vessel location information, including actual distances between vessels, spatial coverage of vessel groups, and vessel density within a vessel group, serving as a key basis for measuring the degree of spatial aggregation of vessels. The preset spatial range threshold is a numerical threshold set by maritime regulators according to the regulatory needs and scenario importance of different sea areas. It is used to determine whether the spatial distribution of vessels meets the sensitive screening standards, thereby achieving differentiated regulation of sea areas. The second sensitive vessel dataset is a vessel dataset that combines the target sensitive features and spatial range requirements, selected by comparing the spatial parameters of the first sensitive vessel dataset with the spatial range threshold. It is a precise candidate set of sensitive vessels with both feature and spatial dimensions. The sensitive vessel features to be determined are the name / behavioral sensitive features possessed by each vessel in the second sensitive vessel dataset, used to calculate the feature matching degree, and are the core basis for quantifying the sensitivity of vessels. Feature matching degree is the degree of matching between the sensitive features of the vessel to be determined and the search conditions. It is determined by the number of successfully matched features and their corresponding weight values. The higher the value, the better the feature matching degree.Spatial clustering is a measure of the degree to which sensitive vessels are clustered in space. It is derived from normalized spatial parameters of the vessels. The higher the clustering and the smaller the spatial coverage, the higher the value of this indicator. The vessel sensitivity intensity value is a quantitative value calculated by integrating feature matching degree and spatial clustering degree. It is the core basis for determining the sensitivity level of a vessel, upgrading the identification of sensitive vessels from qualitative to quantitative assessment. The preset sensitivity intensity level thresholds are thresholds set by maritime regulators according to regulatory priorities, typically divided into four levels: low, medium, high, and extremely high. Different levels correspond to different regulatory handling priorities. Target sensitive vessels are those that have undergone a complete screening process and whose sensitivity intensity value reaches the preset level threshold (generally medium sensitive or above). They are the key targets for maritime regulation. Sensitivity intensity level data is relevant data representing the level of target sensitive vessels. It may include auxiliary information such as feature matching details, real-time location, description of sensitive behavior, and regulatory handling suggestions, providing a comprehensive basis for regulatory decisions.

[0043] In some embodiments of this application, name-sensitive features include sensitive words in a preset sensitive word library and spelling variations of sensitive words. Spelling variations include at least one of character substitution, homophonic rewriting, separator insertion, and abbreviation with missing characters. Character substitution includes substitution of numbers and letters or substitution of special symbols and characters. Behavioral-sensitive features include at least one of the following: abnormal vessel loitering features, abnormal signal features, intrusion into sensitive areas, territorial sea violation features, illegal fishing features, and features of being included in a sanctions list.

[0044] In the above embodiments, the preset sensitive word library refers to a database pre-configured and stored in the system within the maritime regulatory field, containing various illegal, irregular, sensitive, and inappropriate words. It serves as the basic data source for obtaining sensitive words and spelling variations. The library can be dynamically updated according to regulatory needs, covering all sensitive words related to national security, social order, and public order and good morals. Sensitive words refer to the core, unmodified illegal, irregular, sensitive, and inappropriate basic words in the preset sensitive word library. These are core identifiers related to ship names that need to be identified in maritime regulation, such as various explicit sensitive words that violate regulatory provisions. Spelling variations refer to various irregular spellings corresponding to sensitive words in the preset sensitive word library, generated to circumvent conventional sensitive word detection. These are variations of sensitive words, including at least one of character replacement, homophonic rewriting, separator insertion, and abbreviated spelling with missing characters, comprehensively covering the circumvention methods used by violators. Character replacement refers to the method of replacing the original characters in sensitive words with other types of characters. This is a common means for violators to circumvent sensitive word detection, including the substitution of numbers and letters, and the substitution of special symbols and characters. By replacing characters, the appearance of sensitive words is changed while retaining their core meaning. Homophonic rewriting refers to replacing the original characters in sensitive words with homophones or near-homophones, preserving the meaning of the sensitive words through phonetic association and circumventing direct character matching detection. Separator insertion refers to inserting meaningless separators between characters in sensitive words, including spaces, hyphens, and special symbols, thus circumventing detection by splitting the character sequence. Abbreviation and omission refers to simplifying sensitive words by omitting some characters, preserving core identification elements by simplifying the character sequence to circumvent detection. Number and letter substitution refers to replacing letters in sensitive words with numbers of corresponding shapes or pronunciations, or replacing numbers with letters, such as replacing "a" with "4" or "o" with "0," which is a common form of character substitution. The substitution of special symbols and characters refers to replacing ordinary characters in sensitive words with special symbols, or vice versa. For example, replacing "c" with "(" or "s" with "$", etc., to circumvent detection through symbol deformation. Abnormal loitering characteristics refer to the behavior of a vessel in a certain sea area without a reasonable navigation purpose, engaging in prolonged slow navigation, repeated turning back, or lingering. This is determined by analyzing data such as the vessel's navigation trajectory and speed changes. Abnormal signal characteristics refer to behavioral features that do not conform to normal navigation states, such as interrupted signals, false signals, frequent signal switching, and abnormal signal parameters in the Automatic Identification System (AIS), reflecting the potential risk of illegal or irregular operations by the vessel. Intrusion into sensitive areas refers to the behavior of a vessel entering sensitive areas designated by maritime regulations (such as military restricted areas, ecological protection areas, port control areas, etc.) without permission. This is determined by comparing the vessel's position information with the boundary data of the sensitive area.Territorial sea violation refers to the behavior of vessels entering another country's territorial waters without permission. This is determined by comparing vessel location information with territorial sea boundary data and navigation permit documents, and involves regulatory needs related to national sovereignty. Illegal fishing refers to the behavior of vessels fishing in prohibited fishing areas or during prohibited fishing seasons, or using illegal fishing gear or exceeding fishing quotas. This is determined by analyzing data such as vessel operating area, operating time, and equipment information. Being listed on a sanctions list refers to the behavior of vessels being listed on sanctions lists by international organizations, relevant countries, or regions. This is determined by linking to the maritime regulatory sanctions list database and represents a key regulatory characteristic targeting specific violators.

[0045] For name-sensitive features, this includes sensitive words and spelling variations from a pre-defined sensitive word library. The specific types of spelling variations (character replacement, homophonic rewriting, separator insertion, and abbreviation / omission) and the specific forms of character replacement (substitution of numbers and letters, substitution of special symbols and characters) are specified in detail. This ensures that the system can accurately cover various sensitive word transformations used by violators to evade detection when extracting name-sensitive features, avoiding missed detections due to spelling variations. The system first loads the pre-defined sensitive word library, extracts core sensitive words, and then generates various spelling variations through a rule engine to form a complete set of name-sensitive features, ensuring that no possible evasive spellings are missed. For behaviorally sensitive features, it is clearly defined that they include at least one of the following: abnormal loitering features of vessels, abnormal signal features, features of intrusion into sensitive areas, features of territorial waters violations, features of illegal fishing, and features of being included in the sanctions list. This covers sensitive behaviors of vessels in multiple dimensions such as navigation, operation, and identity, so that the system has clear judgment criteria when extracting behaviorally sensitive features. It can comprehensively identify various sensitive behaviors of vessels in navigation and operation, make up for the shortcomings of existing technologies that only focus on name-sensitive features, and realize the upgrade of supervision from "single-dimensional name" to "dual-dimensional name and behavior".

[0046] In one embodiment, a pre-defined sensitive word library contains the sensitive word "luck." System-generated spelling variations include character substitutions such as "l4ck" and "lck," homophonic rewritings such as "luk," separator insertions such as "luck" and "luck," and abbreviated spellings such as "lk" and "luc." When maritime regulatory personnel initiate a search command targeting this sensitive word and related sensitive behaviors, the system, when retrieving the ship name data "luck123," successfully extracts the name's sensitive features using spelling variation recognition rules. Simultaneously, by analyzing the ship's AIS dynamic data, it is found that the ship has been lingering in an ecological protection zone for an extended period (without a legitimate operational purpose), and its sailing speed is abnormally slow with repeated back-and-forth movements. This indicates that the ship possesses two types of behavioral sensitive features: abnormal loitering and intrusion into sensitive areas, thus adding the ship to the sensitive vessel candidate set. By clearly defining the specific types and forms of name and behavioral sensitive features, the system can comprehensively capture various sensitive information about ships. Even if the violator uses evasive spellings or complex combinations of sensitive behaviors, it can still be accurately identified, significantly improving the comprehensiveness and accuracy of sensitive vessel identification. In addition, spelling variants also include substitutions for similar-looking characters, and behaviorally sensitive features include features such as false ship markings and unreported operations, further expanding the feature coverage and improving system adaptability.

[0047] In some embodiments of this application, constructing a sensitive feature dictionary based on name-sensitive features and behavior-sensitive features includes: standardizing or labeling the name-sensitive features and behavior-sensitive features to obtain a standardized sensitive feature set; and constructing a sensitive feature dictionary based on the standardized sensitive feature set.

[0048] In the above embodiments, standardization or tagging processing refers to the normalization and structuring of name-sensitive features and behavior-sensitive features. This aims to eliminate differences in feature formats and distortion interference, ensuring that all features have a unified structure and format to meet the requirements of building a sensitive feature dictionary. Specifically, name-sensitive features undergo standardization operations such as lowercase conversion, character avoidance restoration, and delimiter removal. Behavior-sensitive features are alphabetically encoded according to preset rules to obtain feature tag codes, and corresponding weight values ​​are assigned to different behavior features, with the weight values ​​set according to regulatory priorities. The standardized sensitive feature set refers to the unified structure and standardized format of sensitive features after standardization or tagging. It serves as the direct data foundation for building the sensitive feature dictionary. Each sensitive feature in the set has a unified format and clear attribute identifiers, ensuring the standardization of the dictionary construction and the accuracy of matching.

[0049] First, the name-sensitive features are standardized by performing a lowercase conversion operation to unify all characters into lowercase, eliminating matching errors caused by differences in character case. For example, "Luck" and "LUCK" are both converted to "luck". Then, the avoidance character restoration operation is performed to restore various deformed characters to their original characters. For example, the number "4" is restored to the letter "a" and the special symbol "" is restored to the letter "c". Finally, the separator removal operation is performed to delete meaningless separators, such as spaces, hyphens, and commas, so that "luck" and "luck" are both processed into "luck", thus converting all name-sensitive features into a standard and uniform character sequence. Then, the behaviorally sensitive features are tagged and encoded alphabetically according to preset rules to obtain feature tag codes. For example, abnormal loitering features are coded as "YC", abnormal signal features as "XH", intrusion into sensitive areas as "RQ", territorial waters violation features as "LH", illegal fishing features as "LC", and features listed on the sanctions list as "ZC". Simultaneously, based on the regulatory priority of each type of behaviorally sensitive feature, a corresponding weight value is assigned to each feature tag code. For example, the weight value for territorial waters violation features is set to 0.9, the weight value for illegal fishing features is set to 0.7, and the weight value for abnormal loitering features is set to 0.5, giving the behaviorally sensitive features quantifiable attributes. Through the above standardization or tagging process, a standardized sensitive feature set is obtained, in which all sensitive features have a unified format and clear attribute identifiers. Finally, a sensitive feature trie is constructed based on the standardized sensitive feature set. A multi-branch tree structure is built with standardized single characters (corresponding to name sensitive features) or behavioral feature label codes (corresponding to behavioral sensitive features) as nodes. The path from the root node to the terminal node constitutes a complete sensitive feature string. Sensitive feature strings with a common prefix share the corresponding node. For example, the name sensitive features "luck" and "luckjapan" share the "luck" node sequence, and the behavioral sensitive features "RQ" (intrusion sensitive area) and "RQLC" (intrusion sensitive area + illegal fishing) share the "RQ" node sequence. This significantly saves memory space while ensuring the efficiency and accuracy of the matching process.

[0050] In some embodiments of this application, after constructing a sensitive feature dictionary based on name-sensitive features and behavior-sensitive features, the method includes: configuring a dynamic update triggering mechanism for the sensitive feature dictionary; and performing a dynamic update operation on the sensitive feature dictionary based on the dynamic update triggering mechanism to obtain an updated sensitive feature dictionary.

[0051] In the above embodiments, the dynamic update triggering mechanism refers to the preset conditions used to trigger updates to the sensitive feature dictionary tree. This is the core rule for achieving dynamic maintenance of the dictionary tree, ensuring that it can adapt to changes in regulatory needs and data status in a timely manner. Triggering conditions include adding new sensitive feature types, adjusting regulatory rules, or nodes having expired data updates. Dynamic update operations refer to the maintenance operations performed on the sensitive feature dictionary tree when the dynamic update triggering mechanism is met. This ensures that the dictionary tree always adapts to the latest regulatory needs and vessel data status, including operations such as adding nodes, adjusting weights, or cleaning up invalid nodes, maintaining good retrieval performance and accuracy. The updated sensitive feature dictionary tree refers to the sensitive feature dictionary tree adapted to the latest sensitive feature types, regulatory rules, and vessel data after the dynamic update operation. It has higher adaptability and retrieval accuracy, continuously meeting the dynamic needs of maritime supervision and avoiding missed or false detections caused by a static dictionary tree.

[0052] After constructing a sensitive feature dictionary based on name-sensitive and behavior-sensitive features, the system configures a dynamic update trigger mechanism for this sensitive feature dictionary, specifying the conditions for triggering the dictionary update to ensure the targeted and timely nature of the update operation. When a new sensitive feature type is added to the maritime regulatory field, such as adding behavior-sensitive features like "illegal sand mining" or "false reporting," or new sensitive words and spelling variations, the dynamic update trigger mechanism is triggered. When the rules for judging sensitive features in maritime regulation are adjusted, such as adjusting the judgment criteria for a certain type of behavior-sensitive feature or changing the regulatory priority of sensitive features, the dynamic update trigger mechanism is also triggered. When the data attached to a node in the sensitive feature dictionary exceeds the time limit for updating, such as when the ship data attached to a node exceeds the preset 3-month or 6-month update cycle and no new ship data is added, the dynamic update trigger mechanism is also triggered. When the dynamic update trigger mechanism is triggered, the system performs corresponding dynamic update operations on the sensitive feature trie: For newly added sensitive feature types, a node addition operation is performed, adding a node sequence corresponding to the sensitive feature string to the trie, and configuring feature type identifiers and weight values ​​for terminating nodes to ensure that new features can be included in the trie for retrieval; For changes in feature priority caused by regulatory rule adjustments, a weight adjustment operation is performed, adjusting the weight values ​​of the corresponding terminating nodes to adapt the trie to the new regulatory priority; For cases where node-attached data has expired and not been updated or where nodes have no ship data pointers attached, an invalid node cleanup operation is performed, deleting the invalid node or adding an invalid mark to it, promptly terminating the traversal of invalid branches, improving trie retrieval efficiency, and avoiding invalid nodes consuming system resources.

[0053] In one embodiment, the maritime regulatory authority adds the sensitive feature of "illegal dumping" and its corresponding spelling variants, and simultaneously adds the sensitive feature of "unreported operations." The system detects the new sensitive feature types, triggering a dynamic update mechanism. It then adds nodes to the sensitive feature dictionary, standardizes "illegal dumping" to construct a node sequence, and encodes "unreported operations" as "WB" and constructs a node sequence. It assigns feature type identifiers and weight values ​​("illegal dumping" weight 0.85, "unreported operations" weight 0.75) to the termination nodes of the two new features. After a period of time, the regulatory rules are adjusted, increasing the regulatory priority of the sensitive feature of "territorial sea violation." The system triggers a dynamic update mechanism, performing a weight adjustment operation, changing the weight value of the termination node corresponding to "territorial sea violation" from 0.8 to 0.95. If a node in the dictionary has no actual vessel data attached to its sensitive feature and has not been updated for more than 6 months, the system triggers a dynamic update mechanism, performing an invalid node cleanup operation, deleting the node and its corresponding branches. Through the aforementioned dynamic update process, an updated sensitive feature dictionary is obtained. This dictionary can always adapt to the latest regulatory requirements and vessel data status, avoiding the missed or false detection problems caused by changes in regulatory rules and updates to sensitive features in a static dictionary. Furthermore, the dynamic update triggering mechanism includes conditions such as sensitive feature obsolescence and batch updates of vessel data. Update operations also include node splitting and path optimization, supporting both manual and automatic scheduled updates, further enhancing the flexibility and comprehensiveness of dictionary maintenance.

[0054] In some embodiments of this application, the sensitive feature dictionary is configured with a dynamic update triggering mechanism. The dynamic update triggering mechanism includes adding new sensitive feature types, adjusting regulatory rules, or updating node-attached data that has not been updated within the specified time. The update operations include adding nodes, adjusting weights, or cleaning up invalid nodes.

[0055] In the above embodiments, a new sensitive feature type is added: This refers to the new sensitive feature types that have emerged in the maritime regulatory field due to changes in regulatory needs and the upgrading of illegal methods. These include new name-sensitive feature types (such as new sensitive words and spelling variations) and new behavior-sensitive feature types (such as new types of illegal and irregular navigation operations). These need to be included in the sensitive feature dictionary tree in a timely manner to ensure the comprehensiveness of the screening.

[0056] Regulatory rule adjustments: This refers to the adjustments made by maritime regulatory authorities to the rules for judging sensitive features, regulatory priorities, and screening logic based on changes in national security, social order, and public order and good morals. These adjustments include modifying judgment thresholds, adjusting weight coefficients, and changing feature combination rules. The dictionary tree needs to be adapted synchronously to ensure regulatory accuracy.

[0057] Node-attached data not updated within the specified period: This refers to the ship data pointed to by the ship data pointer attached to the node in the sensitive feature dictionary tree. If the data has not been updated for more than the preset update period (e.g., 3 months or 6 months) and no new ship data has been attached, the sensitive feature corresponding to the node no longer has any practical regulatory significance and needs to be cleaned up to improve retrieval efficiency.

[0058] Node addition: refers to the update operation of adding a node sequence corresponding to the new sensitive feature type in the sensitive feature trie, so that the new sensitive feature can be included in the trie for retrieval. This includes configuring attributes such as feature type identifier and weight value for the new node to ensure that the new feature can be effectively matched.

[0059] Weight adjustment: This refers to adjusting the feature weight value of the terminal node in the sensitive feature trie according to regulatory rules, so as to adapt to the update operation of the new regulatory priority. The weight value directly affects the calculation result of feature matching degree, ensuring that the regulatory focus is consistent with the actual needs.

[0060] Invalid node cleanup: This refers to the deletion of nodes in the sensitive feature trie that have no ship data pointer attached or whose attached data has expired and not been updated, in order to improve the efficiency of trie retrieval. This includes deleting the node itself and its corresponding branch path, reducing invalid traversal, and saving system resources.

[0061] First, the dynamic update trigger mechanism is clearly defined, including three specific conditions: adding new sensitive feature types, adjusting regulatory rules, or exceeding the expiration date for updating node-attached data. This enables the system to clearly identify scenarios requiring updates, avoiding blind and delayed update operations. The trigger condition for adding new sensitive feature types ensures that newly emerging sensitive features are promptly included in the screening scope, addressing the continuous escalation of violations. The trigger condition for adjusting regulatory rules ensures that the dictionary tree can synchronously adapt to changes in regulatory policies, maintaining the accuracy of regulatory direction. The trigger condition for exceeding the expiration date for updating node-attached data ensures a streamlined dictionary tree structure, avoiding invalid nodes consuming system resources and improving retrieval efficiency. Second, the update operations are clearly defined, including three specific forms: adding nodes, adjusting weights, or cleaning up invalid nodes. This enables the system to perform targeted maintenance operations after an update is triggered, ensuring the update effect accurately matches the trigger condition: For adding new sensitive feature types, a node addition operation is performed, constructing the node sequence corresponding to the new sensitive feature in the dictionary tree, configuring feature type identifiers and weight values, ensuring the new feature can be quickly integrated into the existing retrieval system. For adjusting regulatory rules, a weight adjustment operation is performed, adjusting the weight value of the corresponding termination node according to the new regulatory priority, ensuring that the feature matching degree calculation results are consistent with the regulatory focus. If the data attached to a node has expired and not been updated, perform an invalid node cleanup operation, delete the node or add an invalid mark to avoid invalid branches occupying search resources and shorten the matching time.

[0062] In one embodiment, the International Maritime Organization (IMO) adds a batch of sensitive terms related to marine environmental protection. Domestic maritime regulatory authorities simultaneously update their preset sensitive term database. The system identifies the newly added sensitive feature type, triggering a dynamic update mechanism and performing a node addition operation. In the sensitive feature dictionary, a node sequence is constructed for the newly added sensitive terms and their corresponding spelling variants, and a name-sensitive feature identifier and corresponding weight value are configured. If the national maritime regulatory authority issues a new regulatory policy, tightening the criteria for determining the sensitive feature of "intrusion into sensitive areas" (e.g., expanding the scope of sensitive areas) and increasing its regulatory priority, the system detects the regulatory rule adjustment, triggering a dynamic update mechanism and performing a weight adjustment operation, changing the weight value of the termination node corresponding to "intrusion into sensitive areas" from 0.8 to 0.95. If a node in the sensitive feature dictionary corresponds to a sensitive term that has been abandoned by the regulatory authority, and the ship data associated with that node has not been updated for more than 6 months, the system triggers a dynamic update mechanism, performing an invalid node cleanup operation, deleting the node and its corresponding branch path, and the branch will not be traversed in subsequent searches. By clearly defining the dynamic update trigger mechanism and the specific types of update operations, the system's dynamic update process is more standardized and precise, enabling timely responses to changes in various regulatory needs and data status, and ensuring that the sensitive feature trie always maintains efficient and accurate retrieval performance. The dynamic update trigger mechanism also includes conditions such as sensitive feature priority adjustment and batch failure of ship data. Update operations include node merging and feature association optimization, supporting update log recording and rollback operations, facilitating regulatory personnel to trace update history and improving the reliability and traceability of trie maintenance.

[0063] In some embodiments of this application, the feature retrieval conditions include at least one of the following: a single name-sensitive feature, a single behavior-sensitive feature, a logical AND / OR of multiple name-sensitive features, a logical AND / OR of multiple behavior-sensitive features, and a combination match of name-sensitive features and behavior-sensitive features.

[0064] In the above embodiments, a single name sensitive feature refers to a feature retrieval condition that uses only one name sensitive feature as the screening basis. It is one of the basic forms of feature retrieval conditions and is suitable for special screening scenarios targeting specific sensitive words, such as screening only ships whose names contain "Luck".

[0065] Single behavioral sensitive feature: refers to a feature retrieval condition that uses only one behavioral sensitive feature as the screening basis. It is one of the basic forms of feature retrieval conditions and is suitable for special screening scenarios targeting specific sensitive behaviors, such as screening only vessels that engage in illegal fishing.

[0066] Logical AND / OR of multiple name-sensitive features: refers to the feature retrieval conditions formed by combining multiple name-sensitive features as the screening basis through logical AND (simultaneous satisfaction) or logical OR (any one satisfaction). It is applicable to combined name-sensitive feature screening scenarios. The logical AND condition requires that the ship name contains all the specified name-sensitive features at the same time, while the logical OR condition requires that the ship name contains any one of the specified name-sensitive features.

[0067] Logical AND / OR of multiple behaviorally sensitive features: refers to feature retrieval conditions formed by combining multiple behaviorally sensitive features as screening criteria through logical AND (simultaneous satisfaction) or logical OR (any one satisfaction). It is applicable to combined behaviorally sensitive feature screening scenarios. The logical AND condition requires the ship to possess all the specified behaviorally sensitive features simultaneously, while the logical OR condition requires the ship to possess any one of the specified behaviorally sensitive features.

[0068] The combination of name-sensitive features and behavior-sensitive features: This refers to the use of name-sensitive features and behavior-sensitive features as screening criteria, and the feature retrieval conditions formed by logical combination. This can achieve more accurate cross-dimensional screening and is suitable for complex regulatory scenarios. For example, it can screen ships whose names contain "dubo" and have intrusion into sensitive areas, or screen ships whose names contain "japan" or have abnormal signal behavior.

[0069] This application clarifies the specific types of feature retrieval conditions, covering various forms such as single feature retrieval, multi-feature combination retrieval, and cross-dimensional feature combination retrieval. It constructs a flexible and diverse retrieval condition system, enabling maritime regulators to flexibly set screening conditions according to specific regulatory needs, adapting to different regulatory scenarios without modifying the underlying system logic. When the regulatory need is to conduct a special screening targeting a specific sensitive keyword, a single name sensitivity feature can be set as the search condition. The system will only match vessels that possess that name sensitivity feature, achieving precise and targeted supervision. When the regulatory need is to conduct a special screening targeting a specific sensitive behavior, a single behavior sensitivity feature can be set as the search condition. The system will only match vessels that possess that behavior sensitivity feature, focusing on specific violations. When the regulatory need is to screen vessels containing multiple sensitive keywords, a logical AND of multiple name sensitivity features can be set as the search condition. The system will only match vessels that simultaneously meet all name sensitivity features, improving screening accuracy. When the regulatory need is to screen vessels possessing any one of multiple sensitive behaviors, a logical OR of multiple behavior sensitivity features can be set as the search condition, expanding the screening coverage. When the regulatory need is to screen vessels that contain both a specific sensitive name and a specific sensitive behavior, a combination of name sensitivity features and behavior sensitivity features can be set as the search condition, achieving cross-dimensional precise screening and effectively identifying complex violating vessels.

[0070] In some embodiments of this application, the ship space parameters include at least one of the following: actual distance between ships, spatial coverage of the ship group, and ship density of the ship group. The preset spatial range threshold is a graded threshold set by maritime regulators according to the regulatory needs of different sea areas.

[0071] In the above embodiments, the actual distance between ships refers to the straight-line distance between two or more ships, which is one of the core parameters reflecting the spatial distribution of ships. It is calculated using the latitude and longitude coordinates in the ship location information, for example, by using the Haversine formula to calculate the shortest distance between two points on the Earth's surface, and is used to determine whether there is a sensitive situation of close-range ship aggregation. The spatial coverage area of ​​the ship group refers to the geographical area covered by the real-time location information of multiple ships, usually expressed as the area of ​​the smallest convex polygon region containing all ship locations. The unit can be square nautical miles, square kilometers, etc., and is used to determine the concentrated distribution range of the ship group in a certain sea area, adapting to the screening needs of sensitive ship aggregation in local sea areas. Ship density of the ship group refers to the number of ships per unit space area, which is an important parameter reflecting the degree of ship aggregation. It is calculated by dividing the number of ships in the group by the area of ​​the spatial coverage area of ​​the ship group. The unit can be ships / square nautical miles, ships / square kilometer, etc. A higher density indicates a higher degree of ship aggregation and a greater sensitive risk. Tiered thresholds: These refer to the spatial range thresholds set by maritime regulators for different levels based on factors such as the regulatory importance, shipping activity, and sensitivity of different sea areas. For example, the actual distance threshold between vessels in core sensitive sea areas (such as military restricted areas and core areas of ecological protection zones) is set at 500 meters, while the actual distance threshold between vessels in general regulatory sea areas is set at 1 nautical mile, adapting to the differentiated regulatory needs of different levels of sea areas.

[0072] First, the system clarifies that spatial parameters include at least one of the following: actual distance between ships, spatial coverage of a ship group, and ship density within a ship group. This provides a clear basis for calculating spatial parameters, allowing the system to select appropriate parameters for calculation based on specific regulatory needs, comprehensively covering spatial distribution across different dimensions. After extracting ship location information from the first sensitive ship dataset, the system calculates the actual distance between ships using a geographic coordinate distance calculation formula (such as the Haversine formula) to accurately reflect the proximity relationship between ships. When calculating the spatial coverage of a ship group, the system constructs a minimum convex polygon containing all ship locations and calculates the area of ​​this polygon, intuitively reflecting the breadth of the ship group's spatial distribution. When calculating the ship density of a ship group, the system divides the number of ships in the group by the area of ​​the spatial coverage of the group to obtain the number of ships per unit area, quantifying the degree of ship aggregation. Second, the system clarifies that the preset spatial range thresholds are tiered thresholds set by maritime regulators based on the regulatory needs of different sea areas. This makes the threshold settings more targeted, adapting to the regulatory differences in different sea areas and avoiding the problem of insufficient screening accuracy caused by using a uniform threshold. Maritime regulators set differentiated thresholds for different sea areas based on factors such as sensitivity level, shipping activity, and regulatory objectives. For core sensitive sea areas (such as military restricted zones and core areas of ecological protection zones), strict spatial range thresholds are set, such as a 500-meter threshold for actual distance between vessels, a 0.5 square nautical mile threshold for the spatial coverage of a vessel group, and a vessel density threshold of 5 vessels per square nautical mile. For general regulatory sea areas, relatively lenient spatial range thresholds are set, such as a 1-nautical-mile threshold for actual distance between vessels, a 1-square-nautical-mile threshold for the spatial coverage of a vessel group, and a vessel density threshold of 3 vessels per square nautical mile. For busy port areas, thresholds can be adjusted according to actual conditions to balance regulatory efficiency and shipping convenience. During the spatial screening process, the system compares the calculated vessel spatial parameters with the corresponding sea area's thresholds, selecting vessels whose spatial parameters meet the threshold requirements to form a second sensitive vessel dataset, achieving precise screening based on both feature dimensions and spatial dimensions.

[0073] like Figure 2 As shown, Figure 2This is the overall architecture of a sensitive vessel discovery system corresponding to sensitive names with spatial characteristics and vessel discovery methods. It is divided into two core modules: the sensitive vessel discovery function on the left client side and the real-time vessel data processing service on the right server side. These two modules work together to support the intelligent identification and monitoring of sensitive vessels. The server side, as the underlying data support, encompasses four core services: vessel position data access, external interface processing, vessel position data management, and sensitive dictionary vessel tree management. It is responsible for the collection, interaction, and governance of multi-source vessel position data, as well as the construction and maintenance of the sensitive dictionary tree, providing data and algorithmic structural support for the entire system. The client side provides practical functions for regulatory users, including sensitive word and region input, real-time vessel position monitoring, sensitive vessel scanning, sensitive vessel query, sensitive vessel positioning and monitoring, sensitive vessel recording, and sensitive vessel alarms. It realizes the entire business process operation from sensitive rule configuration and real-time scanning and identification to target vessel query and tracking and alarm recording. All functions of the client side are based on the underlying data processing and dictionary tree management capabilities of the server side, forming a complete system closed loop of "back-end data support + front-end business operation".

[0074] like Figure 3 As shown, Figure 3 This is a schematic diagram of ship navigation status and sensitive target monitoring, intuitively demonstrating the application effect of the sensitive ship detection system in actual maritime scenarios, such as... Figure 3 Similar monitoring diagrams can use the sea as a background, with time nodes marked by numbers and core monitoring areas identified by letters, combined with ship icons, text labels, and line symbols to construct a dynamic ship monitoring scenario. Figure 3 The system has two core monitoring areas, A and B. Through visualization, the core functions of the sensitive vessel detection system, namely "area delineation, target identification, trajectory tracking, and multi-target monitoring", are implemented in a real-world scenario. The combination of numerical serial numbers, geometric areas, and vessel markings accurately reflects the complete application process of the system from data collection to situation presentation. It serves as intuitive evidence of the "sensitive vessel positioning and surveillance" and "sensitive area monitoring" functions in the technical solution.

[0075] like Figure 4As shown, an embodiment of this application provides a sensitive vessel determination device 200, including: a first acquisition module 210, a first extraction module 220, a first construction module 230, a second acquisition module 240, a first determination module 250, a second determination module 260, a third determination module 270, a fourth determination module 280, a fifth determination module 290, and a sixth determination module 292. The first acquisition module 210 is used to acquire vessel name data of multiple vessels to be detected in response to a retrieval command for sensitive vessels; the first extraction module 220 is used to extract name-sensitive features and behavior-sensitive features of each vessel to be detected based on the vessel name data; the first construction module 230 is used to construct a sensitive feature dictionary tree based on the name-sensitive features and behavior-sensitive features; and the second acquisition module 240 is used to perform sensitive feature matching with the sensitive feature dictionary tree according to the feature retrieval conditions carried by the retrieval command to acquire a first sensitive vessel dataset. The first determining module 250 is used to extract ship position information from the first sensitive ship dataset and determine ship spatial parameters; the second determining module 260 is used to determine the second sensitive ship dataset based on the ship spatial parameters and a preset spatial range threshold; the third determining module 270 is used to determine the name sensitive features and / or behavior sensitive features of the sensitive ships to be judged based on the second sensitive ship dataset; the fourth determining module 280 is used to determine the feature matching degree based on the feature retrieval conditions, name sensitive features and / or behavior sensitive features; the fifth determining module 290 is used to determine the ship sensitivity intensity value based on the spatial clustering degree and feature matching degree of the second sensitive ship dataset, wherein the spatial clustering degree is used to characterize the degree of clustering of the sensitive ships to be judged; the sixth determining module 292 is used to determine the target sensitive ship based on the ship sensitivity intensity value and a preset sensitivity intensity level threshold, and output the sensitivity intensity level data of the target sensitive ship.

[0076] The device described in this application executes the entire process of "data acquisition, dual feature extraction, trie construction, feature matching, spatial parameter determination, dual screening, matching degree and sensitivity intensity quantification, and target vessel judgment output" through ten modules in a coordinated manner. This enables multi-dimensional and accurate identification, hierarchical judgment and automated screening of sensitive vessels, which greatly improves the efficiency, accuracy and pertinence of maritime supervision.

[0077] like Figure 5 As shown, an embodiment of this application provides a sensitive vessel determination apparatus 300, including a processor 302 and a memory 304. The memory 304 stores a program or instructions. When the processor 302 executes the program or instructions in the memory 304, it implements the steps of the sensitive vessel determination method as described in any of the above embodiments. Therefore, the sensitive vessel determination apparatus 300 possesses all the beneficial effects of the sensitive vessel determination method as described in any of the above embodiments.

[0078] Embodiments of this application provide a readable storage medium storing a program or instructions that, when executed by a processor, implement the steps of the sensitive vessel determination method as described in any of the above embodiments. Therefore, the readable storage medium possesses all the beneficial effects of the sensitive vessel determination method as described in any of the above embodiments.

[0079] In the claims, description, and accompanying drawings of this invention, the term "plural" refers to two or more. Unless otherwise explicitly defined, the terms "upper," "lower," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are used only for the convenience of describing the invention and simplifying the descriptive process, and are not intended to indicate or imply that the device or element referred to must have the described specific orientation, or be constructed and operated in a specific orientation. Therefore, these descriptions should not be construed as limiting the invention. The terms "connected," "installed," "fixed," etc., should be interpreted broadly. For example, "connected" can be a fixed connection between multiple objects, a detachable connection between multiple objects, or an integral connection; it can be a direct connection between multiple objects or an indirect connection between multiple objects through an intermediate medium. For those skilled in the art, the specific meaning of the above terms in this invention can be understood based on the specific circumstances described above.

[0080] In the claims, description, and accompanying drawings of this invention, the terms "one embodiment," "some embodiments," "specific embodiment," etc., refer to a specific feature, structure, material, or characteristic described in connection with that embodiment or example, which is included in at least one embodiment or example of the invention. In the claims, description, and accompanying drawings of this invention, illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0081] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for identifying sensitive vessels, characterized in that, include: In response to a search command for sensitive vessels, the system retrieves the vessel name data of multiple vessels to be detected. Based on the ship name data, extract the name-sensitive features and behavior-sensitive features for each of the ships to be detected; Construct a sensitive feature trie based on the name-sensitive features and the behavior-sensitive features; Based on the feature retrieval conditions carried by the retrieval instruction, sensitive feature matching is performed with the sensitive feature trie to obtain the first sensitive ship dataset; Extract the ship position information from the first sensitive ship dataset to determine the ship spatial parameters; The second sensitive ship dataset is determined based on the ship spatial parameters and the preset spatial range threshold. Based on the second sensitive vessel dataset, determine the name-sensitive features and / or behavioral-sensitive features of the vessels to be identified as sensitive. The feature matching degree is determined based on the feature retrieval conditions, the name-sensitive features, and / or the behavior-sensitive features; Based on the spatial clustering degree of the second sensitive ship dataset and the feature matching degree, the ship sensitivity intensity value is determined, wherein the spatial clustering degree is used to characterize the degree of clustering of the sensitive ships to be determined; Based on the ship's sensitivity intensity value and the preset sensitivity intensity level threshold, the target sensitive ship is determined, and the sensitivity intensity level data of the target sensitive ship is output.

2. The method for determining sensitive vessels according to claim 1, characterized in that, The name sensitivity features include: sensitive words in a preset sensitive word library and spelling variations of the sensitive words. The spelling variations include at least one of character replacement, homophonic rewriting, separator insertion, and abbreviation with missing characters. The character replacement includes: substitution of numbers and letters or substitution of special symbols and characters. The behaviorally sensitive characteristics include at least one of the following: abnormal vessel loitering, abnormal signaling, intrusion into sensitive areas, territorial waters violation, illegal fishing, or being listed on a sanctions list.

3. The method for determining sensitive vessels according to claim 2, characterized in that, The step of constructing a sensitive feature trie based on the name-sensitive features and the behavior-sensitive features includes: The name-sensitive features and behavior-sensitive features are standardized or labeled to obtain a standardized sensitive feature set. Construct a sensitive feature dictionary tree based on the standardized sensitive feature set.

4. The method for determining sensitive vessels according to claim 1, characterized in that, After constructing the sensitive feature trie based on the name-sensitive features and the behavior-sensitive features, the method further includes: Configure a dynamic update trigger mechanism for the sensitive feature trie; Based on the dynamic update triggering mechanism, a dynamic update operation is performed on the sensitive feature dictionary to obtain the updated sensitive feature dictionary.

5. The method for determining sensitive vessels according to claim 4, characterized in that, The sensitive feature dictionary is configured with a dynamic update triggering mechanism. The dynamic update triggering mechanism includes adding new sensitive feature types, adjusting regulatory rules, or updating node-attached data that has not been updated within the specified period. Update operations include adding nodes, adjusting weights, or cleaning up invalid nodes.

6. The method for determining sensitive vessels according to claim 1, characterized in that, The feature retrieval conditions include at least one of the following: a single name-sensitive feature, a single behavior-sensitive feature, a logical AND / OR of multiple name-sensitive features, a logical AND / OR of multiple behavior-sensitive features, and a combination of name-sensitive features and behavior-sensitive features.

7. The method for determining sensitive vessels according to claim 1, characterized in that, The ship space parameters include at least one of the following: actual distance between ships, spatial coverage of the ship group, and ship density of the ship group. The preset spatial range threshold is a graded threshold set by maritime regulators according to the regulatory needs of different sea areas.

8. A device for identifying sensitive ships, characterized in that, include: The first acquisition module is used to acquire the ship name data of multiple ships to be detected in response to the search command for sensitive ships; The first extraction module is used to extract the name-sensitive features and behavior-sensitive features of each of the ships to be detected based on the ship name data; The first construction module is used to construct a sensitive feature trie based on the name sensitive features and the behavior sensitive features; The second acquisition module is used to perform sensitive feature matching with the sensitive feature trie based on the feature retrieval conditions carried by the retrieval instruction to obtain the first sensitive ship dataset. The first determining module is used to extract ship position information from the first sensitive ship dataset and determine ship spatial parameters. The second determining module is used to determine the second sensitive ship dataset based on the ship spatial parameters and the preset spatial range threshold. The third determining module is used to determine the name sensitive features and / or behavior sensitive features of the sensitive vessel to be determined based on the second sensitive vessel dataset; The fourth determining module is used to determine the feature matching degree based on the feature retrieval conditions, the name-sensitive features, and / or the behavior-sensitive features; The fifth determining module is used to determine the sensitivity intensity value of a ship based on the spatial clustering degree of the second sensitive ship dataset and the feature matching degree, wherein the spatial clustering degree is used to characterize the degree of clustering of the sensitive ships to be determined; The sixth determination module is used to determine the target sensitive vessel based on the vessel's sensitivity intensity value and a preset sensitivity intensity level threshold, and output the sensitivity intensity level data of the target sensitive vessel.

9. A device for identifying sensitive ships, characterized in that, include: processor; A memory storing programs or instructions, wherein the processor, when executing the programs or instructions in the memory, implements the steps of the method for determining sensitive ships as described in any one of claims 1 to 7.

10. A readable storage medium, characterized in that, The readable storage medium stores a program or instructions that, when executed by a processor, implement the steps of the method for determining a sensitive vessel as described in any one of claims 1 to 7.

Images