A data retrieval, mining and processing method and system for character database

Abstract

The present application relates to a data retrieval and mining processing method and system for a character database, and relates to the field of data processing technology, including: obtaining a target character and event category and inputting them into a database, retrieving and obtaining related data, type, and update time; analyzing the priority of the data type and arranging them into a first data sequence; adjusting the type priority in combination with the relevant time of the target event category and the data update time, obtaining the adjusted type priority and arranging them into a second sequence; generating a basic weight and a time weight according to the target event category, optimizing the two data sequences, obtaining an optimal display data sequence, and displaying it as a retrieval and mining result. The present invention solves the problem that traditional character data retrieval and display are only sorted by time relevance or type, which is insufficient in applicability and affects the efficiency and accuracy of data resolution and call, and improves the comprehensive applicability of data display and retrieval accuracy.

Description

Technical Field

[0001] The present application relates to the field of data processing, and in particular to a method and system for data retrieval, mining and processing of a character database. Background Art

[0002] With the development of intelligent data applications, efficient retrieval and precise mining of character databases have become a crucial foundation for improving information processing and regulatory efficiency in related fields. Currently, traditional character data retrieval and display rely primarily on time correlation or single-type sorting, requiring manual intervention and adjustment. These methods suffer from issues such as insufficient applicability, low data resolution and retrieval efficiency, and poor accuracy, making them difficult to meet the real-time retrieval needs of character data in complex scenarios.

[0003] The existing data display method relies solely on a single dimension of time or type for sorting, resulting in a large deviation between the display results and actual application needs. This not only increases the risk of misjudgment and manual calibration costs in information screening, but also makes it difficult to meet the requirements for accurate and efficient retrieval of personal data in intelligent transformation. Summary of the Invention

[0004] In order to solve the above technical problems, the present application provides a data retrieval, mining and processing method and system for a character database, which solves the problems of insufficient applicability, low data resolution and retrieval efficiency, and poor accuracy caused by the traditional character data display being only sorted by time correlation or type.

[0005] The embodiments of this application disclose the following technical solutions:

[0006] In a first aspect, an embodiment of the present application provides a data retrieval and mining method for a character database, the method comprising:

[0007] Obtain a target person and a target event category to be searched, input them into a person database, and retrieve multiple associated data, multiple data types, and multiple data update times;

[0008] Analyzing multiple type priorities of multiple data types and arranging them to obtain a first data sequence;

[0009] Analyzing event-related time according to the target event category, adjusting multiple type priorities in combination with the multiple data update times to obtain multiple adjusted type priorities, and arranging them to obtain a second data sequence;

[0010] A basic weight and a time weight are generated according to the target event category, and the display data sequence is optimized based on the first data sequence and the second data sequence to obtain the optimal display data sequence, which is displayed as a retrieval and mining result.

[0011] In a second aspect, an embodiment of the present application provides a data retrieval and mining processing system for a character database, the system comprising:

[0012] The retrieval information acquisition module is used to obtain the target person and target event category to be searched, input the person database, and retrieve multiple related data, multiple data types, and multiple data update times;

[0013] a type priority analysis module, configured to analyze multiple type priorities of multiple data types and arrange them to obtain a first data sequence;

[0014] a priority adjustment and sorting module, configured to analyze event-related time according to the target event category, adjust the priorities of the multiple types in combination with the multiple data update times, obtain multiple adjusted type priorities, and arrange them to obtain a second data sequence;

[0015] The weight optimization display module is used to generate basic weights and time weights according to the target event category, optimize the display data sequence based on the first data sequence and the second data sequence, obtain the optimal display data sequence, and display it as a retrieval and mining result.

[0016] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0017] This application proposes a data retrieval, mining and processing method and system for a character database, which realizes the accurate display of character-related data through multi-dimensional data fusion and dynamic weight optimization. First, the target person and event category are obtained, and the pre-built character database is input to retrieve the related data, and the data type and update time are obtained synchronously. By analyzing the historical call frequency of the data type, the type priority is calculated and sorted to generate the first data sequence; at the same time, combined with the historical average retrieval time of the target event, the similarity between the data update time and the event-related time is calculated, and the type priority is adjusted to obtain the second data sequence. Further, based on the similarity between the mean of the sample event-related time and the current event-related time, the basic weight and time weight are generated, and the two sequences are optimized. The related data are randomly sorted by genetic algorithm, the sorting fitness is calculated and iteratively optimized, and finally converged to obtain the optimal display sequence that takes into account the historical frequency and event time characteristics, and displayed as the retrieval and mining result.

[0018] The technical solution of this application realizes intelligent sorting of character data by integrating multi-dimensional information such as historical call frequency, event time characteristics and dynamic weight parameters, solving the problem of insufficient applicability caused by traditional methods relying only on time or type sorting. It effectively improves the comprehensive applicability and retrieval accuracy of data display in complex scenarios, and provides a reliable solution for intelligent data processing. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0020] Figure 1 A flowchart of a data retrieval and mining method for a character database provided in an embodiment of the present application;

[0021] Figure 2 A schematic diagram of the structure of a data retrieval, mining and processing system for a character database provided in an embodiment of the present application;

[0022] In the accompanying drawings, the components represented by the reference numerals are described as follows:

[0023] Retrieval information acquisition module 01, type priority analysis module 02, priority adjustment and sorting module 03, weight optimization and display module 04. DETAILED DESCRIPTION

[0024] The present application provides a data retrieval, mining and processing method and system for a character database, which is used to solve the technical problems in the prior art that the character data display is only sorted according to time correlation or type, resulting in insufficient applicability, low data resolution and call efficiency, and poor accuracy.

[0025] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without making creative efforts are within the scope of protection of this application.

[0026] In the description of this application, the terms "first" and "second" are used for descriptive purposes only and should not be understood to indicate or imply relative importance or implicitly specify the number of the technical features indicated. Therefore, a feature specified as "first" or "second" may explicitly or implicitly include one or more of the described features. In the description of this application, "plurality" means two or more, unless otherwise specifically specified.

[0027] In the description of this application, the term "for example" is used to mean "used as an example, illustration or explanation". Any embodiment described as "for example" in this application is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is given to enable any person skilled in the art to implement and use the present invention. In the following description, details are listed for the purpose of explanation. It should be understood that a person of ordinary skill in the art will recognize that the present invention can be implemented without using these specific details. In other examples, well-known structures and processes will not be elaborated in detail to avoid obscuring the description of the present invention with unnecessary details. Therefore, the present invention is not intended to be limited to the embodiments shown, but is consistent with the widest scope consistent with the principles and features disclosed in this application.

[0028] Example 1, as shown in the attached Figure 1 As shown, the present application provides a data retrieval and mining processing method for a character database, the method comprising the following steps:

[0029] S100: Obtain a target person and a target event category to be searched, input them into a person database, and retrieve multiple associated data, multiple data types, and multiple data update times;

[0030] In an embodiment of the present application, in a character data retrieval and mining scenario, in order to accurately obtain the associated data of a target character for subsequent analysis, the target character and event category need to be input into a pre-built character database to retrieve the associated data.

[0031] Specifically, when the user triggers a search command, the target person's identity and target event category (such as terrorism investigation, missing persons tracking, etc.) are first obtained, and then the target person is entered into the constructed person database.

[0032] The database collects related data of multiple people within a preset historical time (such as public transportation travel, hotel stays, communication records, etc.), constructs an associated topological structure containing people and related data, and can be updated in real time.

[0033] Furthermore, based on the database, multiple related data related to the target person are retrieved (such as taking the subway 5 hours ago, etc.), and the data type of each related data (such as public transportation, etc.) and the data update time (that is, the time interval from the current time, such as "5 hours ago, etc.) are extracted.

[0034] The associated data, data types and update time obtained through the retrieval will serve as basic information for subsequent data analysis and optimization, ensuring that the retrieval results are more in line with actual application needs.

[0035] Step S100 of the method provided in the embodiment of the present application includes:

[0036] Obtain the target person and target event category to be retrieved;

[0037] Input the target person into a pre-built person database for retrieval to obtain a plurality of related data;

[0038] Get the data type and update time of multiple related data.

[0039] In the embodiment of the present application, in the scenario of character data retrieval and mining, in order to achieve accurate data retrieval and mining, it is necessary to rely on a pre-built character database for retrieval to obtain multiple related data.

[0040] Specifically, based on the actual needs of the relevant affairs field, the identity of the target person (such as name, ID number, etc.) and the target event category (such as terrorism-related investigation, missing person tracking, etc.) are obtained through manual input or system docking.

[0041] Furthermore, the acquired target person information is input into a pre-built person database, and a plurality of associated data of the target person is obtained through database retrieval.

[0042] The steps of constructing the "character database" in the method provided in the embodiment of the present application include:

[0043] Acquire multiple characters and collect related data of the multiple characters within a preset historical time range to obtain multiple related data sets;

[0044] Construct an association topology structure of multiple characters and multiple associated data sets, obtain a character database, and continuously update it.

[0045] In the embodiment of the present application, in order to improve the accuracy and dynamic adaptability of character data retrieval, it is necessary to build a character database covering multi-dimensional related information to provide support for data mining in complex scenarios.

[0046] Specifically, based on the needs of related business fields, we first collect related data of multiple people within a preset time range through multiple source channels (such as public transportation systems, hotel registration platforms, and government databases), covering travel records, accommodation information, communication logs, and other types. After data cleaning and standardization, multiple related data sets are formed.

[0047] Among them, these data sets cover different business scenarios such as terrorism investigation and missing persons tracking, as well as time dimensions such as recent activity trajectories and historical behavior patterns.

[0048] On this basis, graph database technology is used to construct a topological structure of people and related datasets, with nodes representing person entities and edges representing data relationships (such as "taking the subway" and "staying in a hotel"). Data is quantitatively differentiated based on timeliness and business relevance. For example, public transportation records from the past month are prioritized to highlight recent behavior.

[0049] In addition, the database has the ability to update in real time. When new data (such as new hotel check-in records) is generated, it will automatically trigger dynamic adjustment of the topology structure to ensure the real-time nature of data associations, thereby efficiently processing large amounts of character data and providing a dynamic data foundation for subsequent multi-dimensional weight optimization and display sequence generation.

[0050] After the target person is input into the pre-built person database for retrieval, the associated topological structure of multiple persons and associated data sets built in the database will be used to quickly locate and extract multiple associated data related to the target person (covering travel records, accommodation information, communication logs, etc.).

[0051] For example, if the target person is a person under investigation for terrorism-related crimes, after entering their personal information into the personal database, the related data retrieved may include their high-speed rail travel records in the past three months (such as the G123 train travel record from Beijing to Shanghai), the check-in information of a certain hotel chain (including check-in time, co-residents, etc.), and the login log of their mobile phone number in a certain instant messaging software (including communication records such as login device and geographic location).

[0052] After acquiring multiple linked data sets, the data type and update time of each linked data set are also obtained simultaneously. Data types include, but are not limited to, public service data, communication and social data, and government security data. The data update time records the last update time of each data item, providing a basis for subsequent time-based analysis.

[0053] For example, after obtaining the high-speed rail travel records, hotel check-in information, and communication login logs of the same person under investigation for terrorism-related activities, the data types and update times of these related data are simultaneously obtained. Specifically, the high-speed rail travel records (public service data) were updated at 14:30 on June 15, 2025; the hotel check-in information (government security data) was updated at 09:15 on June 10, 2025; and the mobile phone login logs (communication and social media data) were updated at 22:45 on June 20, 2025. This time information provides the basis for subsequent analysis based on the time dimension of terrorism-related incidents.

[0054] S200: Analyze multiple type priorities of multiple data types and arrange them to obtain a first data sequence;

[0055] In the embodiment of the present application, after obtaining multiple associated data and corresponding data types of the target person, in order to achieve orderly display and efficient call of the data, it is necessary to analyze the priority and sort based on the historical call frequency of the data type.

[0056] Specifically, firstly, the number of calls of each data type during the historical time period when searching for person-related data is called is counted.

[0057] Furthermore, the ratio of the number of calls of each data type to the total number of calls is calculated as the type priority.

[0058] Finally, the plurality of associated data are sorted in descending order of type priority to form a first data sequence.

[0059] After this sorting, data with higher type priority is arranged first, providing a basic sorting framework for subsequent in-depth optimization combined with event characteristics.

[0060] Step S200 of the method provided in the embodiment of the present application includes:

[0061] Obtain the number of calls of the multiple data types when performing person-related data retrieval calls within a historical period;

[0062] Calculate the ratio of each call count to the sum of multiple call counts as the type priority, and obtain multiple type priorities;

[0063] The plurality of associated data are sorted in descending order according to the plurality of type priorities to obtain a first data sequence.

[0064] In the embodiment of the present application, after obtaining multiple types of associated data and corresponding data types of the target person, in order to ensure the orderly display of the data, priority analysis and sorting are required based on the historical call frequency of the data type.

[0065] Specifically, we first count the number of calls for each data type during the historical period (e.g., the past year) when searching for person-related data, so as to reflect the degree of attention and frequency of use of different data types in actual applications.

[0066] For example, in a terrorism-related investigation scenario, historical search data shows that the number of calls for public transportation data, accommodation registration data, and communication log data are 200 times, 150 times, and 180 times, respectively, reflecting the usage preference of different data types in case investigation.

[0067] Furthermore, the ratio of each data type's call count to the total number of calls is calculated as the type priority. This ratio can quantitatively represent the relative importance of each data type and provide an objective basis for subsequent sorting.

[0068] Taking the aforementioned data as an example, the total number of calls is 530 (200+150+180). The type priority of public transportation data is 200 / 530≈0.377, hotel check-in data is 150 / 530≈0.283, and communication records is 180 / 530≈0.340, which intuitively reflects the high-frequency characteristics of public transportation data in historical retrieval.

[0069] Finally, the plurality of associated data are sorted in descending order of type priority to form a first data sequence. After this sorting, the associated data corresponding to the data type with the highest priority is arranged first, so that the data type with the highest frequency of use is displayed first.

[0070] For example, in the display of related data for target persons involved in terrorism investigations, public transportation data (type priority 0.377), communication records (type priority 0.340), and hotel check-in data (type priority 0.283) are arranged in sequence, providing a basic sorting framework for subsequent in-depth optimization based on event characteristics, enabling investigators to quickly obtain high-frequency related data and improve the pertinence and efficiency of data retrieval.

[0071] S300: Analyzing event-related time according to the target event category, adjusting multiple type priorities in combination with the multiple data update times to obtain multiple adjusted type priorities, and arranging them to obtain a second data sequence;

[0072] In an embodiment of the present application, after obtaining a first data sequence sorted based on historical call frequency, in order to make the data display more consistent with specific event characteristics, the type priority needs to be dynamically adjusted in combination with the time characteristics of the target event category.

[0073] Specifically, we first analyze the average time that the target event category is called in historical retrieval and use it as the event-related time.

[0074] Furthermore, the similarity between the update time of each associated data and the event-related time is calculated to obtain multiple data confidence coefficients. This coefficient is used to represent the degree of match between the data update time and the event time characteristics. The higher the similarity, the greater the data confidence coefficient.

[0075] Furthermore, the existing type priority is adjusted according to the obtained data confidence coefficient to obtain an adjusted type priority.

[0076] Finally, the associated data is reordered in descending order of adjustment type priority to form a second data sequence. This allows the data sorting to further align with the temporal characteristics of the target event based on historical call frequency, providing a more targeted sorting foundation for subsequent weight optimization and display sequence generation.

[0077] Step S300 in the method provided in the embodiment of the present application includes:

[0078] Obtaining the average time for data retrieval calls for events of the target event category within a historical period as event-related time;

[0079] Calculating similarities between the multiple data update times and the event-related times to obtain multiple data confidence coefficients;

[0080] Adjusting the multiple type priorities according to the multiple data confidence coefficients to obtain multiple adjusted type priorities;

[0081] The plurality of associated data are sorted in descending order according to the plurality of adjustment type priorities to obtain a second data sequence.

[0082] In an embodiment of the present application, after obtaining a first data sequence sorted based on historical call frequency, in order to make the data display more consistent with the time characteristics of a specific event, the type priority needs to be dynamically calibrated in combination with the time characteristics of the target event category.

[0083] Specifically, we first analyze the average time the target event category is retrieved in historical searches and use this as the event relevance time. For example, for a missing person incident that occurred six hours ago, the event relevance time can be determined as six hours; while the event relevance time for a terrorist incident might be the data update period within the past three months.

[0084] Furthermore, the similarity between the update time of each associated data and the event-related time is calculated to generate a data confidence coefficient, which is obtained through the formula "data confidence coefficient (similarity) = 1-|data update time-event-related time| / event-related time". This formula quantifies the timeliness of the data through the relative difference in time intervals.

[0085] For example, if the time between an event and the update time of the associated data is 6 hours, and the update time of the associated data is 5 hours, the similarity is 1-|5-6| / 6≈0.833, which means the data confidence coefficient is 0.833. The closer the data update time is to the event, the higher the data confidence coefficient.

[0086] Furthermore, the existing type priority is adjusted based on the data confidence coefficient to obtain the adjusted type priority, which is obtained through the formula "adjusted type priority = original type priority × data confidence coefficient".

[0087] For example, if the original type priority of a certain data type is 0.3 and its corresponding data confidence coefficient is 0.833, the adjusted type priority is 0.3×0.833≈0.25; if the historical priority of another data type is 0.25 and the data confidence coefficient is 0.9, the adjusted type priority is 0.25×0.9=0.225.

[0088] Finally, the associated data is reordered in descending order of adjustment type priority to form a second data sequence. After this sorting, the associated data corresponding to the data type with the highest adjustment type priority is arranged first, so that the data type with the highest matching degree with the time characteristics of the target event is displayed first.

[0089] For example, the event correlation time for a missing person is 6 hours. The data confidence coefficient (similarity) of the data update time is calculated based on the ratio of the time difference to the event correlation time. For example, if the update time of a communication record is 5 hours and the time difference is 1 hour, its data confidence coefficient is 1-|5-6| / 6≈0.833, and the adjusted type priority after the original type priority of 0.3 is 0.3×0.833≈0.25. For hotel check-in data, the update time is 7 hours and the time difference is 1 hour. Its data confidence coefficient is also 1-|5-6| / 6≈0.833, and the adjusted type priority after the original type priority of 0.283 is 0.283×0.833≈0.236. For public transportation data, the update time is 3 hours and the time difference is 3 hours. Its data confidence coefficient is 1-|3| / 6≈0.5, and the adjusted type priority after the original type priority of 0.377 is 0.377×0.5=0.1885.

[0090] Finally, the adjustment type priorities are ranked from large to small as communication records (0.25), hotel check-in data (0.236), and public transportation data (0.1885), and the corresponding associated data form the second data sequence.

[0091] Based on the first data sequence, this sequence dynamically adjusts data priorities based on the temporal characteristics of the target event category, making it more suitable for the data display needs of specific events. This provides a dual ranking basis for subsequent data analysis based on both historical call frequency and event time characteristics.

[0092] S400: Generate a basic weight and a time weight according to the target event category, optimize the display data sequence based on the first data sequence and the second data sequence, obtain the optimal display data sequence, and display it as a retrieval and mining result.

[0093] In the embodiment of the present application, on the basis of constructing the first data sequence and the second data sequence, in order to achieve the comprehensive applicability of data display, it is necessary to balance the two types of sorting logic through weight generation and sequence optimization.

[0094] Specifically, we first obtain the sample event related time of multiple sample event categories (such as the historical average retrieval time of terrorist-related incidents and missing events), and calculate their average value to obtain the average sample event related time.

[0095] Furthermore, the similarity between the event-related time of the current target event and the average sample event-related time is calculated, and the similarity is multiplied by the preset time weight (such as 0.6) to obtain the time weight; at the same time, the basic weight is obtained by normalizing the overall distribution of the original type priority to ensure that the sum of the two types of weights is 1.

[0096] Furthermore, based on the basic weight and time weight, the first data series and the second data series are weighted and fused for optimization, ultimately obtaining the optimal display data series that takes into account both the historical frequency of the data type and the temporal characteristics of the event.

[0097] This sequence avoids the mechanical nature of relying solely on data type sorting, and overcomes the one-sidedness of sorting based solely on time, making the search and mining results more in line with actual application needs in related transaction scenarios, and improving the comprehensive applicability of data display and retrieval accuracy.

[0098] Step S400 in the method provided in the embodiment of the present application includes:

[0099] Obtaining sample event correlation times of multiple sample event categories and calculating the mean to obtain an average sample event correlation time;

[0100] Calculating the similarity between the event-related time and the average sample event-related time, multiplying the result by a preset time weight to obtain a time weight, and calculating a basic weight;

[0101] According to the basic weight and the time weight, based on the first data sequence and the second data sequence, the display data sequence is optimized to obtain the optimal display data sequence, which is displayed as the retrieval and mining result.

[0102] In an embodiment of the present application, after constructing a first data sequence based on historical call frequency and a second data sequence adjusted in combination with event time characteristics, in order to achieve comprehensive applicability of data display, it is necessary to balance the two types of sorting through weight generation and sequence optimization.

[0103] Specifically, we first obtain the sample event related time of multiple sample event categories (such as the historical average retrieval time of terrorist-related incidents, missing events, and fraud events), and calculate their average value to obtain the average sample event related time.

[0104] For example, the sample events include terrorism (3 months), disappearance (24 hours), and fraud (72 hours), so the average sample event correlation time is (90+1+3) / 3≈31.33 days.

[0105] Furthermore, the similarity between the event-related time of the current target event and the average sample event-related time is calculated, that is, it is obtained through the formula "similarity = 1-|target event-related time-average sample event-related time| / average sample event-related time".

[0106] Furthermore, the similarity is multiplied by a preset time weight (such as 0.6) to obtain the time weight (time weight = similarity × preset time weight); at the same time, the basic weight is obtained by calculating "1-time weight" to ensure that the sum of the two types of weights is 1.

[0107] Among them, the time weight represents the degree of influence of the time characteristics of the target event in the weight distribution. The larger the value, the more similar the time characteristics of the target event are to the time characteristics of the overall sample events, and the more the data display relies on the second data series adjusted based on the event time.

[0108] Similarly, the basic weight indicates the degree of influence of the historical call frequency of the data type in the weight distribution. The larger the value, the greater the difference between the time characteristics of the target event and the time characteristics of the overall sample events, and the more the data display relies on the first data sequence based on the historical call frequency.

[0109] For example, the target event is a disappearance event 6 hours ago, and its event-related time is 0.25 days (6 / 24=0.25). The similarity with the average sample event-related time of 31.33 days is 1-|0.25-31.33| / 31.33≈0.008. Then its time weight is 0.008×0.6≈0.005, and the basic weight is 1-0.005=0.995, indicating that the time characteristics of the target event are quite different from those of the overall sample events. Therefore, the time weight will decrease accordingly, and the basic weight will increase accordingly.

[0110] Furthermore, based on the obtained basic weight and time weight, the first data sequence and the second data sequence are optimized for display data sequence to obtain an optimal display data sequence.

[0111] In the method provided in the embodiment of the present application, the step of “optimizing the display data sequence based on the first data sequence and the second data sequence according to the basic weight and the time weight to obtain the optimal display data sequence” includes:

[0112] Randomly sort multiple related data to obtain a random data sequence;

[0113] Analyzing the sorting fitness of the random data sequence based on the first data sequence, the second data sequence, the basic weight, and the time weight;

[0114] Continue to perform random sorting iterative optimization and converge to obtain the optimal display data sequence with the largest sorting fitness.

[0115] In an embodiment of the present application, in a character data retrieval and mining scenario, in order to achieve optimal data display that takes into account both historical call frequency and event time characteristics, it is necessary to dynamically adjust the data sorting through iterative optimization.

[0116] First, the associated data are randomly sorted to generate an initial random data sequence, which will serve as the basic sample for subsequent optimization.

[0117] Furthermore, based on the first data sequence (historical call frequency sorting), the second data sequence (event time feature adjustment sorting), the basic weight (historical frequency influence) and the time weight (event time influence), the sorting fitness of the random sequence is calculated.

[0118] In the method provided in the embodiment of the present application, the step of “analyzing the sorting fitness of the random data sequence based on the first data sequence, the second data sequence, the basic weight, and the time weight” includes:

[0119] Calculating similarities between the random data sequence and the first data sequence and the second data sequence respectively to obtain a first random similarity and a second random similarity;

[0120] The similarity between the first random similarity and the basic weight is calculated, and the similarity between the second random similarity and the time weight is calculated, and the average is calculated to obtain the sorting fitness.

[0121] In the embodiment of the present application, in the scenario of character data retrieval and mining, in order to quantitatively evaluate the comprehensive rationality of random data sequences, it is necessary to integrate the dual dimensions of historical call frequency and event time characteristics.

[0122] Specifically, the similarities between the random sequence and the first and second data sequences are calculated respectively, that is, the first random similarity and the second random similarity are obtained by calculating the proportion of common elements or the similarity of the sequence position.

[0123] The similarity between the random sequence and the first and second data sequences is calculated by counting the number of associated data in the same position between the random sequence and the first data sequence, and dividing the result by the total number of data to obtain the first random similarity; similarly, the second random similarity is calculated by counting the number of associated data in the same position between the random sequence and the second data sequence, and dividing the result by the total number of data.

[0124] For example, assume the total number of data is 5, the first data sequence is [A, B, C, D, E], and the random sequence is [B, A, C, E, D]. In this case, if the associated data C is in the same position in the first data sequence and the random sequence, then there is only one associated data C in the same position, and the first random similarity is 1 / 5 = 0.2. If the second data sequence is [C, A, B, E, D], and the associated data A, E, and D are in the same position in the second data sequence and the random sequence, then there are three associated data A, E, and D in the same position in the random sequence and the second data sequence, and the second random similarity is 3 / 5 = 0.6.

[0125] In another embodiment, the similarity between the random sequence and the first and second data sequences can be calculated using the rank similarity of the associated data. Specifically, for each associated data item, its rank in the random sequence and the target data sequence is calculated. The rank similarity is calculated as 1 minus the ratio of the absolute difference between the two data items to their mean (i.e., similarity = 1 - |rank difference| / rank mean).

[0126] For example, suppose there is a set of related data. A certain related data is ranked 5 in the random sequence and 7 in the first data sequence. According to the similarity calculation formula, the positional similarity of this related data is 1-|5-7| / [(5+7) / 2] = 1-2 / 6 ≈ 0.67. The positional similarities of all related data are averaged to obtain the first random similarity and the second random similarity.

[0127] The first random similarity is the average of the similarities between the random sequence and all the associated data in the first data sequence, and the second random similarity is the average of the similarities between the random sequence and all the associated data in the second data sequence.

[0128] Furthermore, the similarity between the first random similarity and the basic weight, and the similarity between the second random similarity and the time weight are calculated, and the two are added together to take the average as the sorting fitness, that is, it is obtained through the formula "sorting fitness = (similarity between the first random similarity and the basic weight + similarity between the second random similarity and the time weight) / 2". The higher the value, the stronger the comprehensive rationality of the sequence.

[0129] The similarity between the first random similarity and the basic weight is calculated using the formula "similarity = 1 - |first random similarity - basic weight| / [(first random similarity + basic weight) / 2]". This formula quantifies the degree of matching by calculating the relative proportion of the difference between the two.

[0130] Similarly, the similarity between the second random similarity and the time weight is calculated as "similarity = 1-|second random similarity-time weight| / [(second random similarity+time weight) / 2]". Finally, the sorting fitness is determined by the mean of the two to comprehensively evaluate the rationality of the comprehensive matching of the random sequence and the weight system.

[0131] For example, in a terrorism investigation scenario, assuming the first random similarity is 0.6 and the second random similarity is 0.4, if the base weight is 0.7 and the time weight is 0.3, the similarity between the first random similarity and the base weight is calculated as 1-|0.6-0.7| / [(0.6+0.7) / 2]=1-0.1 / 0.65≈0.85, and the similarity between the second random similarity and the time weight is 1-|0.4-0.3| / [(0.4+0.3) / 2]=1-0.1 / 0.35≈0.71. Adding the two and taking the average gives a ranking fitness of (0.85+0.71) / 2≈0.78.

[0132] This value indicates that the random sequence has a higher degree of match with the historical call frequency sorting (first data sequence), but a relatively low degree of match with the event time feature sorting (second data sequence). It is necessary to further adjust the data position through iterative optimization to improve the overall rationality.

[0133] Furthermore, a genetic algorithm is used to iteratively optimize the data, and the current random sequence is subjected to selection, crossover and mutation operations to generate a new candidate sequence.

[0134] Specifically, the algorithm evaluates the quality of each candidate sequence by ranking its fitness, retaining the sequences with the highest fitness as the parent sequence, and generating offspring sequences through crossover (e.g., partial match crossover) and mutation (e.g., positional mutation). If the offspring sequence has a higher fitness than the parent, the parent is replaced; otherwise, the offspring is retained with a certain probability to avoid falling into a local optimum.

[0135] For example, in a terrorism investigation scenario, the initial random sequence had a sorting fitness of 0.27. Through genetic algorithm iteration, a communication record (with a high degree of match between update time and event time characteristics) was moved from the 5th position to the 2nd position. The number of positions shared with the first data sequence remained at 3 (first random similarity 0.6), while the number of positions shared with the second data sequence increased to 3 (second random similarity 0.6).

[0136] Furthermore, the similarity between the first random similarity and the basic weight of 0.7 is calculated to be 1-|0.6-0.7| / [(0.6+0.7) / 2]≈0.85, and the similarity between the second random similarity and the time weight of 0.3 is 1-|0.6-0.3| / [(0.6+0.3) / 2]≈1-0.3 / 0.45≈0.33. The new sorting fitness is (0.85+0.33) / 2≈0.59.

[0137] Furthermore, after 1,000 iterations, the fitness converged to 0.55. At this time, the first three data in the sequence simultaneously met the historical high-frequency calls (such as high-speed rail travel records) and event time matching (such as communication logs within 24 hours), forming an optimal display sequence that takes into account the dual dimensions of historical frequency and event time, and displayed it as a retrieval mining result.

[0138] The genetic algorithm optimization process dynamically balances the influence of basic weights and time weights, and continuously evolves the sequence using selection, crossover, and mutation operations. The final displayed data sequence ensures the priority of high-frequency data types in historical retrieval and reflects the timeliness of current event time data, thereby achieving comprehensive applicability and improved retrieval accuracy in related transaction scenarios.

[0139] The embodiments of the present application achieve the following technical effects through the above specific implementation methods:

[0140] The embodiment of the present application provides a data retrieval and mining processing method for a character database. First, the target character and event category to be retrieved are obtained, and a pre-built character database is input to retrieve the associated data, data type and update time. The database collects multi-source data to construct an associated topological structure and continuously updates it to ensure the comprehensiveness and timeliness of the data. Secondly, the historical call frequency of the data type is analyzed, the proportion of the call times is calculated as the type priority, and the first data sequence is obtained by sorting from large to small; at the same time, the historical average retrieval time of the target event is combined to calculate the similarity between the data update time and the event-related time to obtain the data confidence coefficient, and the type priority is adjusted accordingly to obtain the second data sequence. Then, the mean of the sample event-related time is obtained, the similarity between the current event and the mean is calculated to generate the basic weight and time weight, and the display data sequence is optimized based on the two sequences. Finally, the associated data is randomly sorted, the similarity with the two sequences is calculated and the sorting fitness is obtained in combination with the weights, and the optimal display sequence is obtained as the retrieval and mining result through iterative optimization convergence.

[0141] The method provided in the embodiment of the present application solves the problems of insufficient applicability and poor accuracy caused by traditional character data retrieval relying only on a single sorting of time or type. Through the steps of "data retrieval-priority analysis-weight generation-sequence optimization", the comprehensive applicability of data display to complex scenarios and the retrieval accuracy are improved, providing a scientific solution for the intelligent processing of character data in related business fields.

[0142] Example 2, as shown in the attached Figure 2 As shown, based on the inventive concept of a data retrieval, mining and processing method for a character database provided in Example 1, the present application also provides a data retrieval, mining and processing system for a character database, specifically comprising:

[0143] The search information acquisition module 01 is used to obtain the target person and target event category to be searched, input the person database, and retrieve multiple related data, multiple data types, and multiple data update times;

[0144] Type priority analysis module 02, used to analyze multiple type priorities of multiple data types and arrange them to obtain a first data sequence;

[0145] Priority adjustment and sorting module 03, configured to analyze event-related time according to target event categories, adjust multiple type priorities in combination with the multiple data update times, obtain multiple adjusted type priorities, and arrange them to obtain a second data sequence;

[0146] The weight optimization display module 04 is used to generate basic weights and time weights according to the target event category, optimize the display data sequence based on the first data sequence and the second data sequence, obtain the optimal display data sequence, and display it as a retrieval and mining result.

[0147] In one embodiment, the retrieval information acquisition module 01 is further configured to:

[0148] Obtain the target person and target event category to be retrieved;

[0149] Input the target person into a pre-built person database for retrieval to obtain a plurality of related data;

[0150] Get the data type and update time of multiple related data.

[0151] In one embodiment, the type priority analysis module 02 is further configured to:

[0152] Obtain the number of calls of the multiple data types when performing person-related data retrieval calls within a historical period;

[0153] Calculate the ratio of each call count to the sum of multiple call counts as the type priority, and obtain multiple type priorities;

[0154] The plurality of associated data are sorted in descending order according to the plurality of type priorities to obtain a first data sequence.

[0155] In one embodiment, the priority adjustment and sorting module 03 is further configured to:

[0156] Obtaining the average time for data retrieval calls for events of the target event category within a historical period as event-related time;

[0157] Calculating similarities between the multiple data update times and the event-related times to obtain multiple data confidence coefficients;

[0158] Adjusting the multiple type priorities according to the multiple data confidence coefficients to obtain multiple adjusted type priorities;

[0159] The plurality of associated data are sorted in descending order according to the plurality of adjustment type priorities to obtain a second data sequence.

[0160] In one embodiment, the weight optimization display module 04 is further configured to:

[0161] Obtaining sample event correlation times of multiple sample event categories and calculating the mean to obtain an average sample event correlation time;

[0162] Calculating the similarity between the event-related time and the average sample event-related time, multiplying the result by a preset time weight to obtain a time weight, and calculating a basic weight;

[0163] According to the basic weight and the time weight, based on the first data sequence and the second data sequence, the display data sequence is optimized to obtain the optimal display data sequence, which is displayed as the retrieval and mining result.

[0164] It should be noted that the order in which the embodiments of the present application are presented is for illustrative purposes only and does not necessarily represent the superiority or inferiority of the embodiments. Furthermore, the foregoing descriptions of specific embodiments of this specification are provided. Furthermore, the processes depicted in the accompanying drawings do not necessarily require the specific order or sequential sequence shown to achieve the desired results. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0165] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present application should be included in the scope of protection of the present application.

[0166] This specification and drawings are merely illustrative of the present application and are intended to cover any and all modifications, variations, combinations, or equivalents within the scope of this application. Obviously, those skilled in the art may make various modifications and variations to this application without departing from the scope of this application. Thus, this application is intended to include such modifications and variations as fall within the scope of this application and its equivalents.

Claims

1. A data retrieval and mining method for a character database, characterized in that: The method comprises: Obtain a target person and a target event category to be searched, input them into a person database, and retrieve multiple associated data, multiple data types, and multiple data update times; Analyzing multiple type priorities of multiple data types and arranging them to obtain a first data sequence; Analyzing event-related time according to the target event category, adjusting multiple type priorities in combination with the multiple data update times to obtain multiple adjusted type priorities, and arranging them to obtain a second data sequence, including: Obtaining the average time for data retrieval calls for events of the target event category within a historical period as event-related time; Calculating similarities between the multiple data update times and the event-related times to obtain multiple data confidence coefficients; Adjusting the multiple type priorities according to the multiple data confidence coefficients to obtain multiple adjusted type priorities; sorting the plurality of associated data in descending order according to the plurality of adjustment type priorities to obtain a second data sequence; A basic weight and a time weight are generated according to the target event category, and the display data sequence is optimized based on the first data sequence and the second data sequence to obtain the optimal display data sequence, which is displayed as a retrieval and mining result.

2. The data retrieval and mining processing method for a character database according to claim 1, characterized in that: Obtain the target person and target event category to be searched, input the person database, and retrieve multiple related data, multiple data types, and multiple data update times, including: Obtain the target person and target event category to be retrieved; Input the target person into a pre-built person database for retrieval to obtain a plurality of related data; Get the data type and update time of multiple related data.

3. The data retrieval and mining processing method for a character database according to claim 2, characterized in that: The steps of constructing the character database include: Acquire multiple characters and collect related data of the multiple characters within a preset historical time range to obtain multiple related data sets; Construct an association topology structure of multiple characters and multiple associated data sets, obtain a character database, and continuously update it.

4. The data retrieval and mining processing method for a character database according to claim 1, characterized in that: Analyzing multiple type priorities of multiple data types and arranging them to obtain a first data sequence includes: Obtain the number of calls of the multiple data types when performing person-related data retrieval calls within a historical period; Calculate the ratio of each call count to the sum of multiple call counts as the type priority, and obtain multiple type priorities; The plurality of associated data are sorted in descending order according to the plurality of type priorities to obtain a first data sequence.

5. The data retrieval and mining processing method for a character database according to claim 1, characterized in that: Generating a basic weight and a time weight according to the target event category, optimizing the display data sequence based on the first data sequence and the second data sequence, obtaining an optimal display data sequence, and displaying the optimal display data sequence as a search and mining result, including: Obtaining sample event correlation times of multiple sample event categories and calculating the mean to obtain an average sample event correlation time; Calculating the similarity between the event-related time and the average sample event-related time, multiplying the result by a preset time weight to obtain a time weight, and calculating a basic weight; According to the basic weight and the time weight, based on the first data sequence and the second data sequence, the display data sequence is optimized to obtain the optimal display data sequence, which is displayed as the retrieval and mining result.

6. The data retrieval and mining processing method for a character database according to claim 5, characterized in that: Optimizing the display data sequence based on the first data sequence and the second data sequence according to the basic weight and the time weight to obtain an optimal display data sequence includes: Randomly sort multiple related data to obtain a random data sequence; Analyzing the sorting fitness of the random data sequence based on the first data sequence, the second data sequence, the basic weight, and the time weight; Continue to perform random sorting iterative optimization and converge to obtain the optimal display data sequence with the largest sorting fitness.

7. The data retrieval and mining processing method for a character database according to claim 6, characterized in that: Analyzing the sorting fitness of the random data sequence based on the first data sequence, the second data sequence, the basic weight, and the time weight includes: Calculating similarities between the random data sequence and the first data sequence and the second data sequence respectively to obtain a first random similarity and a second random similarity; The similarity between the first random similarity and the basic weight is calculated, and the similarity between the second random similarity and the time weight is calculated, and the average is calculated to obtain the sorting fitness.

8. A data retrieval and mining processing system for a character database, characterized in that: The system is used to execute the data retrieval and mining processing method for a character database according to any one of claims 1 to 7, and the system includes: The retrieval information acquisition module is used to obtain the target person and target event category to be searched, input the person database, and retrieve multiple related data, multiple data types, and multiple data update times; a type priority analysis module, configured to analyze multiple type priorities of multiple data types and arrange them to obtain a first data sequence; The priority adjustment and sorting module is used to analyze the event-related time according to the target event category, adjust the priorities of multiple types in combination with the multiple data update times, obtain multiple adjusted type priorities, and arrange them to obtain a second data sequence, including: Obtaining the average time for data retrieval calls for events of the target event category within a historical period as event-related time; Calculating similarities between the multiple data update times and the event-related times to obtain multiple data confidence coefficients; Adjusting the multiple type priorities according to the multiple data confidence coefficients to obtain multiple adjusted type priorities; sorting the plurality of associated data in descending order according to the plurality of adjustment type priorities to obtain a second data sequence; The weight optimization display module is used to generate basic weights and time weights according to the target event category, optimize the display data sequence based on the first data sequence and the second data sequence, obtain the optimal display data sequence, and display it as a retrieval and mining result.

Images