Computer resource adjustment method and device, electronic equipment and computer readable medium

Abstract

Embodiments of the present disclosure disclose a computer resource adjustment method and device, electronic equipment and a computer readable medium. A specific embodiment of the method comprises: in response to receiving data acquisition information sent by a terminal, generating each structured information data; performing data detection processing on the information data source to obtain a data detection result corresponding to the information data source; determining each target structured information data; performing data extraction processing on the target structured information data to obtain structured extraction data corresponding to the target structured information data; generating target structured extraction data; generating to-be-processed structured transfer information and to-be-processed structured negative association information; and based on the to-be-processed structured transfer information and the to-be-processed structured negative association information, performing a corresponding computer resource adjustment task. The embodiment can improve data processing efficiency and reduce waste of computing resources.

Description

Technical Field

[0001] Embodiments of this disclosure relate to the field of computer technology, and more specifically to computer resource adjustment methods, apparatus, electronic devices, and computer-readable media. Background Technology

[0002] Computer resource adjustment based on data source detection involves adjusting resource allocation during data extraction and processing by performing resource checks on the extracted information. Currently, the common approach to adjusting computer resources is as follows: based on data request information sent by the server, data related to the requested information is retrieved from various data sources. Then, key information is extracted from the retrieved data. Finally, resource detection is performed on the extracted key information, and resource allocation is adjusted based on the detection results.

[0003] However, when adjusting computer resources using the above methods, the following technical problems often arise:

[0004] Extracting key information from data requires extracting all the acquired data, which can lead to low data processing efficiency. Furthermore, directly extracting key information from the acquired data can result in a high degree of overlap with previously extracted key information, leading to a waste of computational resources. Summary of the Invention

[0005] The summary portion of this disclosure is intended to provide a brief overview of the concepts, which will be described in detail in the detailed description portion. This summary portion is not intended to identify key or essential features of the claimed technical solutions, nor is it intended to limit the scope of the claimed technical solutions.

[0006] Some embodiments of this disclosure provide computer resource adjustment methods, apparatuses, electronic devices, and computer-readable media based on data source detection to address one or more of the technical problems mentioned in the background section above.

[0007] In a first aspect, some embodiments of this disclosure provide a computer resource adjustment method based on data source detection. The method includes: in response to receiving data acquisition information sent by a terminal, generating various structured information data, wherein each structured information data corresponds to an information data source, and each structured information data is data organized according to a predetermined structure; for each information data source corresponding to the structured information data, performing data detection processing on the information data source to obtain a data detection result corresponding to the information data source; determining various target structured information data based on the obtained data detection results; and for each target structured information data, adjusting the structured information data... The target structured information data is processed by data extraction to obtain structured extracted data corresponding to the target structured information data. This structured extracted data includes structured transfer information and structured negative correlation information. Based on the obtained structured extracted data, target structured extracted data is generated, including target structured transfer information and target structured negative correlation information. Based on the target structured extracted data, unprocessed structured transfer information and unprocessed structured negative correlation information are generated. The unprocessed structured transfer information includes various structured transfer fields, and the unprocessed structured negative correlation information includes various structured negative correlation fields. Based on the unprocessed structured transfer information and the unprocessed structured negative correlation information, corresponding computer resource adjustment tasks are executed.

[0008] Secondly, some embodiments of this disclosure provide a data push device, comprising: a first generation unit configured to generate various structured information data in response to receiving data acquisition information sent by a terminal, wherein each structured information data corresponds to an information data source, and each structured information data is data organized according to a predetermined structure; a detection unit configured to perform data detection processing on each information data source corresponding to the various structured information data to obtain a data detection result corresponding to the information data source; a determination unit configured to determine various target structured information data based on the obtained data detection results; and a data extraction unit configured to extract data from each target structured information data. The structured information data undergoes data extraction processing to obtain structured extracted data corresponding to the aforementioned target structured information data. The structured extracted data includes structured transfer information and structured negative correlation information. A second generation unit is configured to generate target structured extracted data based on the obtained structured extracted data. The target structured extracted data includes target structured transfer information and target structured negative correlation information. A third generation unit is configured to generate unprocessed structured transfer information and unprocessed structured negative correlation information based on the aforementioned target structured extracted data. The unprocessed structured transfer information includes various structured transfer fields, and the unprocessed structured negative correlation information includes various structured negative correlation fields. An execution unit is configured to execute corresponding computer resource adjustment tasks based on the unprocessed structured transfer information and the unprocessed structured negative correlation information.

[0009] Thirdly, some embodiments of this disclosure provide an electronic device, including: one or more processors; and a storage device having one or more programs stored thereon, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement the method described in any implementation of the first aspect above.

[0010] Fourthly, some embodiments of this disclosure provide a computer-readable medium having a computer program stored thereon, wherein the program, when executed by a processor, implements the method described in any of the implementations of the first aspect above.

[0011] The above embodiments of this disclosure have the following beneficial effects: the computer resource adjustment method based on data source detection in some embodiments of this disclosure can improve data processing efficiency and reduce computational resource waste. Specifically, the reasons for low data processing efficiency and wasted computational resources are: when extracting key information from data, all acquired data needs to be extracted, which easily leads to low data processing efficiency. Moreover, when directly extracting key information from the acquired data, the extracted key information is prone to high repetition with previously extracted key information, resulting in wasted computational resources during extraction. Based on this, the computer resource adjustment method based on data source detection in some embodiments of this disclosure first generates various structured information data in response to receiving data acquisition information sent by the terminal. Each of the structured information data corresponds to an information data source, and each of the structured information data is data organized according to a predetermined structure. Thus, the original data to be processed can be obtained, and the acquired data can be converted into structured data through specific data acquisition tools, which can improve the efficiency of subsequent data processing. Secondly, for each information data source corresponding to the aforementioned structured information data, data detection processing is performed on the aforementioned information data source to obtain the corresponding data detection result. This allows detection of whether the data source has been updated since the last data acquisition. Then, based on the obtained data detection results, each target structured information data is determined. Therefore, only updated data can be processed. Next, for each target structured information data, data extraction processing is performed on the aforementioned target structured information data to obtain the corresponding structured extracted data, which includes structured transfer information and structured negative correlation information. Then, based on the obtained structured extracted data, target structured extracted data is generated, which includes target structured transfer information and target structured negative correlation information. This allows for the integration of the various structured extracted data. Then, based on the aforementioned target structured data extraction, unprocessed structured transfer information and unprocessed structured negative correlation information are generated. The unprocessed structured transfer information includes each structured transfer field, and the unprocessed structured negative correlation information includes each structured negative correlation field. Therefore, technical personnel can further inspect the data to remove duplicate and redundant information, improving the accuracy of the final generated data. Finally, based on the unprocessed structured transfer information and the unprocessed structured negative correlation information, corresponding computer resource adjustment tasks are executed. This allows the data to be pushed to the target server.Furthermore, after data acquisition, the data crawled from various data sources can be directly converted into structured data using a data acquisition device, and then processed, thus improving data processing efficiency. Also, before data extraction, the update time of each data source can be detected, and the results can be used to filter out data sources that have been updated. This allows for the selection of updated data from the acquired data, and only the updated data is processed, reducing the probability of duplicate data being processed repeatedly and minimizing the waste of computational resources during data processing. Attached Figure Description

[0012] The above and other features, advantages, and aspects of the embodiments of this disclosure will become more apparent from the accompanying drawings and the following detailed description. Throughout the drawings, the same or similar reference numerals denote the same or similar elements. It should be understood that the drawings are schematic, and elements are not necessarily drawn to scale.

[0013] Figure 1 These are flowcharts of some embodiments of the computer resource adjustment method based on data source detection according to this disclosure;

[0014] Figure 2 This is a schematic diagram of the structure of some embodiments of the computer resource adjustment apparatus based on data source detection according to the present disclosure;

[0015] Figure 3 This is a schematic diagram of the structure of an electronic device suitable for implementing some embodiments of the present disclosure. Detailed Implementation

[0016] Embodiments of this disclosure will now be described in more detail with reference to the accompanying drawings. While some embodiments of this disclosure are shown in the drawings, it should be understood that this disclosure can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of this disclosure. It should be understood that the accompanying drawings and embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of protection of this disclosure.

[0017] It should also be noted that, for ease of description, only the parts relevant to the invention are shown in the accompanying drawings. Unless otherwise specified, the embodiments and features described in this disclosure can be combined with each other.

[0018] It should be noted that the concepts of "first" and "second" mentioned in this disclosure are used only to distinguish different devices, modules or units, and are not used to limit the order of functions performed by these devices, modules or units or their interdependencies.

[0019] It should be noted that the terms "a" and "a plurality of" used in this disclosure are illustrative rather than restrictive, and those skilled in the art should understand that, unless otherwise expressly indicated in the context, they should be understood as "one or more".

[0020] The names of messages or information exchanged between multiple devices in the embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of such messages or information.

[0021] This disclosure will now be described in detail with reference to the accompanying drawings and embodiments.

[0022] Figure 1 A flow 100 of some embodiments of a computer resource adjustment method based on data source detection according to the present disclosure is shown. This computer resource adjustment method based on data source detection includes the following steps:

[0023] Step 101: In response to receiving data acquisition information sent by the terminal, generate various structured information data.

[0024] In some embodiments, the executing entity (e.g., a computing device) of the computer resource adjustment method based on data source detection can generate various structured information data in response to receiving data acquisition information sent by a terminal. The terminal can be a pre-set timer or a target terminal. The timer can be a device capable of sending preset information at pre-set intervals. The target terminal can be a terminal device corresponding to a technician. The data acquisition information can be information used to prompt the executing entity to begin acquiring structured information data. Each structured information data can be structured data related to value data (e.g., bond transaction data). Each structured information data is data organized according to a predetermined structure. The predetermined structure can be a format for organizing, managing, and storing data. For example, the predetermined structure can be a table structure. Each structured information data can include various fields. Each field corresponds to a group of field values. Each structured information data corresponds to an information data source. The information data source can be the data source corresponding to the structured information data. For example, the information data source can be the URL where the structured information data is located. The aforementioned information data sources have corresponding data update times. These data update times represent the time of the most recent update of the aforementioned information data sources. The aforementioned executing entity can be a client capable of acquiring the aforementioned structured information data and processing it. For example, the aforementioned executing entity can be a computing device capable of data processing and displaying the data.

[0025] In practice, the aforementioned implementing entity can obtain various structured information data from a pre-set structured information database. This structured information database can be a pre-set database used to store the various structured information data.

[0026] In the process of adopting technical solutions to address the aforementioned technical problems, the following issues often arise:

[0027] When acquiring a large amount of data, directly extracting key information from it can lead to low processing efficiency. Conversely, when acquiring a small amount of data and allocating significant storage resources, a large portion of these resources may remain idle and unusable by other processes, resulting in wasted storage resources.

[0028] In response to the aforementioned technical problems, the following solution was adopted:

[0029] Optionally, after step 101, the aforementioned executing entity may also perform the following steps:

[0030] The first step involves performing resource detection processing on each of the aforementioned structured information data sets to obtain memory resource usage data corresponding to each set. This memory resource usage data represents the size of memory occupied by each of the structured information data sets. In practice, the executing entity can use a memory detection function to perform resource detection processing on each of the structured information data sets to obtain memory resource usage data. This memory detection function can be any function capable of detecting the memory size corresponding to the data. For example, the memory detection function could be the `sys.getsizeof()` function.

[0031] The second step, in response to determining that the above memory resource usage data is greater than the preset memory allocation data, is to perform the following steps:

[0032] The first sub-step involves executing an alarm notification task corresponding to the aforementioned memory resource usage data. The memory allocation data can be a pre-allocated memory capacity for storing the various structured information data. For example, the memory allocation data could be 8GB. In practice, the executing entity can emit vibrations to execute the alarm notification task corresponding to the aforementioned memory resource usage data.

[0033] The second sub-step involves displaying a pre-defined prompt window. This prompt window can be a pop-up window used to confirm whether the user needs to adjust resource allocation. The prompt window may include a prompt message, a confirmation control, and a cancellation control. The prompt message can be information used to confirm with the user whether resource allocation needs to be adjusted. For example, the prompt message could be, "The current amount of data being processed is large; do you need to adjust resource allocation?" The confirmation control can be a control indicating that resource allocation adjustment is needed. The cancellation control can be a control indicating that resource allocation adjustment is not needed.

[0034] The third sub-step, in response to detecting a selection operation applied to the confirmation controls included in the aforementioned prompt window, executes the following steps:

[0035] Sub-step one involves increasing the physical memory corresponding to each of the aforementioned structured information data sets to adjust the storage resources associated with them. The selection operation can include, but is not limited to, at least one of the following: clicking, hovering, swiping, and dragging. In practice, firstly, the executing entity can determine the amount of memory to be allocated as the difference between the memory resource usage data and the memory allocation data. Then, the executing entity can use a memory call function to allocate the memory corresponding to the amount to be allocated to each of the aforementioned structured information data sets, thereby increasing the physical memory corresponding to each of the structured information data sets. The memory call function can be a function capable of adjusting memory allocation. For example, the memory call function can be the malloc function.

[0036] Sub-step two involves adding processing threads corresponding to each of the aforementioned structured information data sets to adjust the processor resources for each set. These processing threads can be the threads used during data processing. The processor resources can be the computing resources of the processor. For example, the processor resources can be the number of allocated processing threads. In practice, the executing entity can add processing threads corresponding to each of the aforementioned structured information data sets through a thread adjustment interface to adjust the processor resources for each set of structured information data sets. This thread adjustment interface can be an interface capable of adjusting the number of threads used during data processing. For example, the thread adjustment interface can be an ExecutorService.

[0037] Third, in response to the determination that the memory resource usage data is less than the memory allocation data, the physical memory corresponding to each of the structured information data is reduced to adjust the storage resources corresponding to each of the structured information data. In practice, the executing entity can use the memory call function to allocate idle memory allocated to each of the structured information data to other processes for use, thereby adjusting the storage resources corresponding to each of the structured information data.

[0038] The above technical solution and its related content, combined with steps 101 to 107, constitute an inventive point of this disclosure, solving the problem of "wasted storage resources." Factors leading to wasted computing resources often include: when a large amount of data is acquired, directly extracting key information from the acquired data can easily lead to low processing efficiency during data processing. When a small amount of data is acquired and a large amount of storage resources are allocated to the data, a large amount of storage resources remain idle but cannot be used by other processes, easily causing storage resource waste. Solving these factors can reduce storage resource waste. To achieve this effect, this disclosure first performs resource detection processing on each of the above-mentioned structured information data to obtain memory resource usage data corresponding to each of the above-mentioned structured information data. Thus, the memory size occupied by each structured information data can be obtained. Second, in response to determining that the above-mentioned memory resource usage data is greater than a preset memory allocation data, the following steps are executed: Then, based on the above-mentioned memory resource usage data, an alarm prompt task corresponding to the above-mentioned memory resource usage data is executed. Then, a preset prompt window is displayed, wherein the prompt window includes a confirmation control and a cancellation control. Next, in response to detecting a selection operation on the confirmation control included in the aforementioned prompt window, the following steps are performed: First, the physical memory corresponding to each of the aforementioned structured information data is increased to adjust the storage resources corresponding to each of the aforementioned structured information data. Thus, when the memory allocated to each of the structured information data is less than the actual memory required by each of the structured information data, the memory allocated to each of the aforementioned structured information data can be increased. Then, the processing threads corresponding to each of the aforementioned structured information data are increased to adjust the processor resources corresponding to each of the aforementioned structured information data. Thus, multiple threads can be used to process each of the structured information data, thereby improving data processing efficiency. Finally, in response to determining that the memory resource usage data is less than the memory allocation data, the physical memory corresponding to each of the aforementioned structured information data is reduced to adjust the storage resources corresponding to each of the aforementioned structured information data. Thus, when the actual memory required by each of the structured information data is less than the allocated memory, idle memory can be released, thereby reducing storage resource waste. Also, because the actual memory size used by each of the structured information data can be determined first, and then the memory allocated to each of the structured information data is dynamically adjusted, storage resource waste can be reduced. Furthermore, when it is detected that the memory occupied by each structured information data is large, the number of processing threads for processing the aforementioned structured information data can be increased, thereby improving the processing efficiency of data processing.

[0039] Step 102: For each information data source in each information data source corresponding to each structured information data, perform data detection processing on the information data source to obtain the data detection result of the corresponding information data source.

[0040] In some embodiments, the execution entity may perform data detection processing on each information data source corresponding to each of the structured information data to obtain the data detection result corresponding to the information data source.

[0041] In some optional implementations of certain embodiments, the aforementioned execution entity may perform data detection processing on the aforementioned information data source through the following steps to obtain the data detection result corresponding to the aforementioned information data source:

[0042] The first step involves generating an update comparison result based on the aforementioned information data source, pre-stored historical data update times, and the aforementioned data update time. Each historical data update time corresponds to the last time the structured information data was acquired. Each historical data update time corresponds to an information data source. The update comparison result can be information obtained from the aforementioned data update time, indicating whether the information data source has been updated. For example, the update comparison result can be "data updated" or "data not updated". In practice, firstly, the executing entity can determine the historical data update time corresponding to the aforementioned information data source from the aforementioned historical data update times as the data update time to be compared. Then, a comparison operator can be used to compare the data update time to be compared with the aforementioned data update time to obtain a comparison logical value. The comparison operator can be the "=" operator. The comparison logical value can be a character indicating whether the data update time to be compared is the same as the aforementioned data update time. For example, the comparison logical value can be "True" or "False". The "True" value indicates that the update time of the data to be compared is the same as the update time of the data to be compared. The "False" value indicates that the update time of the data to be compared is different from the update time of the data to be compared. Then, in response to determining the comparison logic value as "True", "data not updated" can be determined as the updated comparison result. In response to determining the comparison logic value as "False", "data updated" can be determined as the updated comparison result.

[0043] The second step involves generating update flag data corresponding to the structured information data source mentioned above, based on the structured information data source itself. This update flag data can be the hash value corresponding to the structured information data. In practice, the executing entity can generate the hash value corresponding to the structured information data as the update flag data using a hash algorithm. This hash algorithm can be the MD5 algorithm (Message-Digest Algorithm 5).

[0044] The third step involves generating a flag data comparison result based on the preset historical update flag data and the aforementioned update flag data. Each historical update flag data can be the update flag data corresponding to the information data source when data was last retrieved from the information data source. Each historical update flag data corresponds to an information data source. The flag data comparison result can be information obtained from the aforementioned update flag data, used to characterize whether the aforementioned information data source has been updated. For example, the flag data comparison result can be "data updated" or "data not updated". In practice, firstly, the executing entity can determine the historical update flag data corresponding to the aforementioned information data source from the aforementioned historical update flag data as the update flag data to be compared. Then, the comparison operator can be used to compare the update flag data to be compared with the aforementioned update flag data to obtain comparison logic data. The comparison logic data can be characters used to characterize whether the update flag data to be compared and the aforementioned update flag data are the same. For example, the comparison logic data can be "True" or "False". Then, in response to determining that the above comparison logic data is "True", "Data not updated" can be determined as the above flag data comparison result. In response to determining that the above comparison logic data is "False", "Data updated" can be determined as the above flag data comparison result.

[0045] The fourth step involves generating a data detection result based on the aforementioned update comparison results and the aforementioned flag data comparison results. This data detection result can be information used to characterize whether the aforementioned information data source has been updated. For example, the data detection result can be "data updated" or "data not updated." In practice, in response to determining that both the update comparison results and the aforementioned flag data comparison results are "data updated," the executing entity can determine "data updated" as the data detection result. In response to determining that either the update comparison results or the aforementioned flag data comparison results contain "data not updated," "data not updated" can be determined as the data detection result.

[0046] Step 103: Based on the obtained data detection results, determine the structured information data of each target.

[0047] In some embodiments, the executing entity can determine each target structured information data based on the obtained data detection results. Each target structured information data can be structured information data that has been updated.

[0048] In some optional implementations of certain embodiments, the aforementioned execution entity may determine each target structured information data based on the obtained data detection results through the following steps:

[0049] The first step is to identify the target detection results based on the obtained data detection results, specifically those that meet the preset data detection conditions. The data detection condition can be that the data detection result is "data has been updated".

[0050] The second step involves identifying the structured information data corresponding to each target detection result as the target's structured information data. Each target structured information data set can include various fields, and each field corresponds to a set of field values.

[0051] Step 104: For each target structured information data in each target structured information data, perform data extraction processing on the target structured information data to obtain the structured extracted data of the corresponding target structured information data.

[0052] In some embodiments, the executing entity may perform data extraction processing on each of the target structured information data to obtain structured extracted data corresponding to the target structured information data. The structured extracted data may be data extracted from the target structured information data. The structured extracted data may include structured transfer information and structured negative correlation information.

[0053] In some optional implementations of certain embodiments, the aforementioned execution entity may perform data extraction processing on the aforementioned target structured information data through the following steps to obtain structured extracted data corresponding to the aforementioned target structured information data:

[0054] The first step involves partitioning the fields of the target structured information data to obtain partitioned field groups. Each partitioned field in each of these groups can be a field from the target structured information data. Each partitioned field in each group corresponds to a set of field values. In practice, the executing entity can group a predetermined number of fields into partitioned field groups to obtain the partitioned field groups. The predetermined number can be a pre-set value. The specific setting of this predetermined number is not limited here.

[0055] The second step involves performing the following data processing steps for each of the aforementioned partitioning field groups:

[0056] The first sub-step involves performing missing value detection processing on each field value group corresponding to the divided field groups, obtaining each missing value detection result group for each of the aforementioned field value groups. Each missing value detection result in each of these result groups can be information used to characterize whether the corresponding field value is missing. For example, the missing value detection result can be "missing" or "not missing". In practice, for each field value in each of the aforementioned field value groups, firstly, in response to determining that the field value is empty, the executing entity can determine "missing" as the missing value detection result corresponding to the field value. In response to determining that the field value is not empty, "not missing" can be determined as the missing value detection result corresponding to the field value.

[0057] The second sub-step involves, for each missing value detection result in each missing value detection result group within the aforementioned missing value detection result groups, performing the following steps in response to determining that the missing value detection result meets the preset missing condition:

[0058] Sub-step one: Determine the field value corresponding to the missing value detection result above as the target field value. Here, the missing condition can be that the missing value detection result is "missing".

[0059] Sub-step two: In response to determining that the field corresponding to the missing value detection result meets the preset field conditions, the preset missing value is identified as the target field value in the partitioned field group, thereby updating the partitioned field group. Here, the aforementioned field conditions can be that the field value group corresponding to the aforementioned field contains numbers. The aforementioned missing value can be a pre-set numerical value. The specific setting of the aforementioned missing value is not limited here.

[0060] Sub-step three: In response to determining that the missing value detection result does not meet the above field conditions, the following steps are performed:

[0061] First, based on the missing value detection results and pre-stored historical target structured information data, the historical field values ​​corresponding to the missing value detection results are determined. The historical target structured information data can be previously stored target structured information data. This historical target structured information data can include various fields. Each field corresponds to a group of field values. The historical field values ​​can be field values ​​obtained from the historical target structured information data that correspond to the missing value detection results. In practice, firstly, the executing entity can determine the field corresponding to the missing value detection results as the field to be processed. Secondly, the field in the historical target structured information data that is identical to the field to be processed can be determined as the target historical field. Then, any field value from the group of field values ​​corresponding to the target historical field can be determined as the historical field value.

[0062] Then, the aforementioned historical field values ​​are determined as the target field values ​​in the partition field group in order to update the partition field group.

[0063] The third sub-step involves identifying each partitioning field in the updated partitioning field group that meets the preset numeric field conditions as a field to be detected. The numeric field condition can be that each field value in the corresponding field value group of the partitioning field contains a number. Each field to be detected corresponds to a field value group. Each field value in the field value group can be the field value corresponding to the field to be detected. Each field value in the field value group corresponds to a unit. The unit can be the unit corresponding to the data. For example, the unit can include, but is not limited to, year, month, and day.

[0064] The fourth sub-step involves standardizing the data of each field value group corresponding to the aforementioned fields to be tested, based on a pre-defined field unit information table, to obtain standardized field value groups. Each standardized field value in each standardized field value group can be a standardized field value. Each standardized field value group corresponds to a standardized field. The standardized field can be the field corresponding to the standardized field value. The field unit information table can be a table representing the correspondence between each unit field and each unit. The field unit information table can include various unit fields. Each unit field corresponds to a unit. Each unit field can be a field whose corresponding field value contains a number. For example, the field unit information table can include, but is not limited to: {Unit field: Remaining term, unit: hours; Unit field: Issuance volume, unit: ten thousand}. In practice, for each field to be tested, the executing entity can first determine the unit field corresponding to the field to be tested from the various unit fields included in the field unit information table as the target unit field. Secondly, the unit corresponding to the aforementioned target unit field can be determined as the target unit for the aforementioned field to be detected. Then, for each field value in the group of field values ​​included in the aforementioned field to be detected, the unit corresponding to the aforementioned field value can be determined as the unit to be converted. Then, the aforementioned unit to be converted, the aforementioned target unit, and the aforementioned field value to be detected can be input into a pre-set data conversion engine to perform data conversion on the aforementioned field value to change the unit of the aforementioned field value from the unit to be converted to the aforementioned target unit, and obtain the data output by the aforementioned data conversion engine as the standardized field value corresponding to the aforementioned field value to be detected. The aforementioned data conversion engine can be a rule engine pre-set by technicians that can perform unit conversion on numbers. For example, when the field value to be detected is 1000, and the corresponding unit to be converted is meters, and the corresponding target unit is kilometers, the aforementioned data conversion engine can convert the field value to be detected 1000 into a value in kilometers, that is, the aforementioned data conversion engine can output 1 as the standardized field value.

[0065] The fifth sub-step involves performing anomaly detection processing on each of the aforementioned standardized field value groups based on a preset field range information group, resulting in anomaly detection result groups for each of the aforementioned standardized field value groups. Each field range information in the aforementioned field range information group can be an interval representing the value range of the corresponding standardized field value. For example, the aforementioned field range information can be [1500, 2000]. Each field range information in the aforementioned field range information group corresponds to a field. Each anomaly detection result in the aforementioned anomaly detection result group can be information representing whether the corresponding standardized field value is abnormal. For example, each anomaly detection result in the aforementioned anomaly detection result group can be "data normal" or "data abnormal".

[0066] In practice, for each of the aforementioned standardized field value groups, the executing entity first determines the standardized field corresponding to the standardized field value group as the target standardized field. Then, it determines the fields in the field unit information group that are identical to the target standardized field as target range fields. Next, it determines the field range information corresponding to the target range field as the field range information to be processed. Then, for each standardized field value in the aforementioned standardized field value group, in response to determining that the standardized field value is within the range represented by the field range information to be processed, "data normal" can be determined as the abnormal field detection result corresponding to the standardized field value. In response to determining that the standardized field value is not within the range represented by the field range information to be processed, "data abnormal" can be determined as the abnormal field detection result corresponding to the standardized field value. Finally, the obtained abnormal field detection results can be combined into an abnormal field detection result group.

[0067] The sixth sub-step involves performing data repair processing on the standardized field value groups based on the aforementioned anomaly field detection result groups, resulting in repaired field value groups. Each repaired field value in each repaired field value group can be a standardized field value that has undergone data repair processing. Each repaired field value group corresponds to a repaired field. The repaired field can be the field corresponding to the repaired field value group.

[0068] The seventh sub-step involves combining the repair fields corresponding to the above repair field value groups into a repair field group.

[0069] The third step involves generating structured transfer information and structured negative correlation information based on the preset structured transfer information fields, the preset negative correlation information fields, and the obtained repair field groups. Each of the aforementioned structured transfer information fields can be a field related to the structured transfer information. Each of the aforementioned negative correlation information fields can be a field corresponding to the structured negative correlation information. The aforementioned structured transfer information can be information related to the circulation of goods. The aforementioned structured transfer information can include, but is not limited to: data source, item name (e.g., security name), item value (e.g., security face value), item circulation volume (e.g., security issuance volume), issuance start date, and issuance end date. For example, each of the aforementioned structured transfer information fields can be, but is not limited to: data source, item name (e.g., security name), item value (e.g., security face value), item circulation volume (e.g., security issuance volume), issuance start date, and issuance end date. The aforementioned structured negative correlation information can be information related to the item value attribute. The aforementioned structured negative correlation information may include, but is not limited to: item acquisition information (e.g., credit entity issuing securities), item attributes (e.g., principal), and value-related information (e.g., interest). For example, each of the aforementioned negative correlation information fields may be, but is not limited to, item acquisition information (e.g., credit entity issuing securities), item attributes (e.g., principal), and value-related information (e.g., interest).

[0070] In practice, the aforementioned executing entity can identify each repair field in each repair field group that is identical to any one of the structured transfer information fields in the aforementioned structured transfer information fields as a separate transfer repair field. Then, for each of these transfer repair fields, firstly, the corresponding repair field value can be identified as the transfer repair field value. Next, the transfer repair field and its value can be combined to form value transfer sub-information. Finally, the combined value transfer sub-information can be combined to form structured transfer information.

[0071] In practice, the aforementioned executing entity can identify each repair field in each repair field group that is identical to any one of the negative association information fields in the aforementioned negative association information fields as a negative association repair field. Then, for each of the aforementioned negative association repair fields, firstly, the repair field value corresponding to the aforementioned negative association repair field can be identified as the negative association repair field value. Then, the aforementioned negative association repair field and the aforementioned negative association repair field value can be combined into value negative association sub-information. Finally, the combined value negative association sub-information can be combined into structured negative association information.

[0072] The fourth step is to combine the above-mentioned structured transfer information and the above-mentioned structured negative correlation information into structured extraction data corresponding to the above-mentioned target structured information data.

[0073] In some optional implementations of certain embodiments, the aforementioned execution entity may perform data repair processing on the aforementioned standardized field value groups based on the aforementioned anomaly field detection result groups through the following steps to obtain the respective repaired field value groups:

[0074] The first step is to perform the following steps for each anomaly detection result group in the above anomaly detection result groups:

[0075] The first sub-step involves, for each anomaly detection result in the above anomaly detection result group, performing the following steps in response to determining that the above anomaly detection result meets the preset anomaly field conditions:

[0076] Sub-step one involves determining the standardized field values ​​corresponding to the above-mentioned anomaly field detection results as the target standardized field values. The anomaly field condition can be that the anomaly field detection result is "data anomaly".

[0077] Sub-step two: Determine the field range information corresponding to the above-mentioned abnormal field detection results in the above field range information group as the target field range information.

[0078] Sub-step three: Based on the aforementioned target field range information, determine the average value of the corresponding field within that target field range information. This average value can be the average value corresponding to the aforementioned target field range information. In practice, firstly, the executing entity can determine the maximum value within the interval represented by the aforementioned target field range information as the target maximum value. Secondly, it can determine the minimum value within the interval represented by the aforementioned target field range information as the target minimum value. Then, the average of the aforementioned target maximum value and the aforementioned target minimum value can be determined as the average value of the field.

[0079] Sub-step four involves determining the average value of the above fields as the target standardized field value, and then updating the target standardized field value.

[0080] The second sub-step involves determining the updated target standardized field values ​​as the target standardized field value group corresponding to the above-mentioned anomaly field detection result group.

[0081] The third sub-step involves identifying the aforementioned target standardized field value group as the repair field value group. In practice, the executing entity can also identify the standardized fields corresponding to the aforementioned target standardized field value group as the repair fields.

[0082] In the process of adopting technical solutions to address the aforementioned technical problems, the following issues often arise:

[0083] Extracting key information directly from all the acquired data can easily lead to a large amount of data to be processed, resulting in high data processing complexity and high computational resource consumption.

[0084] In response to the aforementioned technical problems, the following solution was adopted:

[0085] In some optional implementations of certain embodiments, the aforementioned execution entity may perform data extraction processing on the aforementioned target structured information data through the following steps to obtain structured extracted data corresponding to the aforementioned target structured information data:

[0086] The first step, based on the aforementioned target structured information data, is to determine the text information data and the numerical information data. Specifically, the text information data can be data obtained from various text fields and their corresponding text field value groups. Each text field can be a field in the target structured information data whose corresponding field value group consists entirely of text. Each text field value group can be the field value corresponding to a text field. Similarly, the numerical information data can be data obtained from various numeric fields and their corresponding numeric field value groups. Each numeric field can be a field in the target structured information data whose corresponding field value group contains numbers.

[0087] In practice, firstly, all fields in the target structured information data whose corresponding field value groups are all text can be identified as text value fields. Secondly, for each of these text value fields, the corresponding field value group can be identified as a text value field value group. Then, the text value fields and their value groups can be combined into text information sub-data. Finally, the combined text information sub-data can be combined into the final text information data.

[0088] In practice, firstly, the executing entity can identify each field containing a number in the corresponding field value group within the various fields of the target structured information data as a data value field. Secondly, for each of these digital value fields, the corresponding field value group can be identified as a digital value field value group. Then, the digital value fields and their respective groups can be combined into digital information sub-data. Finally, the combined digital information sub-data can be combined into the final digital information data.

[0089] The second step involves feature extraction processing of the aforementioned text information data to obtain the corresponding text embedding information. This text embedding information can be the feature vector corresponding to the text information data. In practice, the executing entity can input the text information data into a pre-trained feature extraction model to obtain the corresponding feature vector as the text embedding information. This feature extraction model can be a neural network that takes text information data or digital information data as input and outputs the corresponding feature vector. For example, the feature extraction model can be a pre-trained recurrent neural network. Pre-training can be a process of fine-tuning the recurrent neural network using a cross-entropy loss function and a training dataset. Each training data point in the training dataset can be text information data or digital information data used for model training.

[0090] The third step involves performing feature extraction processing on the aforementioned digital information data to obtain the corresponding digital embedding information. This digital embedding information can be the feature vector corresponding to the digital information data. In practice, the executing entity can input the aforementioned digital information data into the aforementioned feature extraction model to obtain the corresponding feature vector as the digital embedding information.

[0091] The fourth step involves performing feature mapping processing on the aforementioned text embedding information and number embedding information to obtain text mapping information and number mapping information. The text mapping information can be the mapped text embedding information, and the number mapping information can be the mapped number embedding information. In practice, the executing entity can input the text embedding information into a pre-trained linear layer to obtain the text mapping information. Similarly, it can input the number embedding information into the linear layer to obtain the number mapping information. The linear layer can be a linear layer capable of mapping feature vectors to a preset feature dimension. This feature dimension can be 2-dimensional.

[0092] The fifth step involves performing feature concatenation on the aforementioned text mapping information and number mapping information to obtain hybrid feature information. This hybrid feature information can be a feature vector obtained by fusing the text mapping information and the number mapping information. In practice, the executing entity can use a concatenation function to perform feature concatenation on the text mapping information and the number mapping information to obtain hybrid feature information. This concatenation function can be any function capable of concatenating two feature vectors. For example, the concatenation function could be the `torch.cat` function.

[0093] Step 6: Based on the aforementioned mixed feature information, generate missing mask data corresponding to the mixed feature information. The missing mask data can be a matrix composed of 0s and 1s obtained after processing the mixed feature information. Each column of the missing mask data can correspond to a field in the target structured information data. Each element value in each column of the missing mask data can correspond to a group of field values. Each element in the missing mask data can be used to represent whether the corresponding field value is missing. 0 in the missing mask data represents a missing field value. 1 in the missing mask data represents a non-missing field value. In practice, the executing entity can input the mixed feature information into a missing value detection function and obtain the return value of the missing value detection function as a missing value detection matrix. The missing value detection function can be a function capable of detecting whether data is null. For example, the missing value detection function can be the torch.isnan() function. The missing value detection matrix can be a matrix composed of True and False values ​​obtained after processing the mixed feature information. Each element in the missing value detection matrix can be used to represent whether the corresponding field value is missing. In the missing data detection matrix described above, "True" indicates that the corresponding field value is not missing, and "False" indicates that the corresponding field value is missing. This missing data detection matrix can then be input into a type conversion function to convert "True" to 1 and "False" to 0, resulting in a converted missing data detection matrix used as the missing data mask. The type conversion function can be any function capable of converting Boolean data to integer data. For example, the type conversion function could be the `int()` function.

[0094] Step 7: Based on the missing mask data described above, generate various missing state arrays. Each of these missing state arrays can be an array composed of a column of elements from the missing mask data. In practice, for each column of elements in the missing mask data, the execution entity can combine these elements into an array as the missing state array, based on the row number corresponding to each element in that column, and in the order of the row numbers.

[0095] Step 8: For each of the missing state arrays mentioned above, perform the following steps:

[0096] The first sub-step is to determine the above missing state array as the target missing state array.

[0097] The second sub-step involves identifying each missing state array that meets a preset processing condition from the aforementioned missing state arrays as a separate missing state array to be processed. The processing condition can be that the missing state array is not the target missing state array.

[0098] The third sub-step generates various differential data and distribution data based on the target missing state array and the various missing state arrays to be processed. Each differential data point can be a numerical value characterizing the independence between the target missing state array and the missing state arrays to be processed. Each differential data point corresponds to one missing state array to be processed. Each distribution data point can be a P-value between the target missing state array and the missing state arrays to be processed. Each distribution data point corresponds to one missing state array to be processed.

[0099] In practice, for each of the aforementioned missing state arrays to be processed, firstly, for each preset bit in the preset bit sequence, the executing entity can determine the element in the target missing state array corresponding to the preset bit as the target missing element. Each preset bit in the preset bit sequence can be a pre-set value corresponding to the number of elements in the target missing state array. For example, the preset bit sequence can be {1, 2, 3, 4}. As an example, when the preset bit is 1, the element in the target missing state array corresponding to the preset bit is the first element in the target missing state array. Then, the element in the missing state array to be processed corresponding to the preset bit can be determined as the missing element to be processed. Then, in response to determining that both the target missing element and the missing element to be processed are 1, the sum of the first truth data and a preset value can be determined as the first truth data to update the first truth data. The first truth data can be a value representing the number of times the target missing element and the missing element to be processed are both 1. The preset value can be 1. Then, in response to determining that the target missing element is 0 and the pending missing element is 1, the sum of the second truth data and the preset value can be used as the second truth data to update the second truth data. The second truth data can be a value representing the number of times the target missing element is 0 and the pending missing element is 1 occurs. Then, in response to determining that the target missing element is 1 and the pending missing element is 0, the sum of the third truth data and the preset value can be used as the third truth data to update the third truth data. The third truth data can be a value representing the number of times the target missing element is 1 and the pending missing element is 0 occurs. Then, in response to determining that both the target missing element and the pending missing element are 0, the sum of the fourth truth data and the preset value can be used as the fourth truth data to update the fourth truth data. The fourth truth data can be a value representing the number of times both the target missing element and the pending missing element are 0 occur. Before updating the first, second, third, and fourth true values, all four values ​​can be 0. Then, the updated first true value is used as the element at position (1,1), the updated second true value as the element at position (1,2), the updated third true value as the element at position (2,1), and the updated fourth true value as the element at position (2,2), thus constructing a frequency matrix. This frequency matrix and a preset parameter adjustment expression are then input into the test function to obtain the variance and distribution data.The aforementioned test function can be the `chi2_contingency` function. The aforementioned parameter adjustment expression can be used to adjust the `lambda_` parameter in the aforementioned test function, enabling the test function to perform a G-test on the input data. For example, the aforementioned parameter adjustment expression can be "lambda_ = log-likelihood ratio".

[0100] The fourth sub-step involves generating a weighted data set based on the aforementioned difference data and distribution data. Each weight data point in this set can be a numerical value representing the weight corresponding to the difference data and the distribution data.

[0101] In practice, firstly, the executing entity can determine the negative logarithm corresponding to each of the aforementioned distribution data as negative logarithmic distribution data. Secondly, for each of the determined negative logarithmic distribution data, the difference data corresponding to the aforementioned negative logarithmic distribution data among the aforementioned difference data can be determined as target difference data. Then, the product of the aforementioned target difference data and the aforementioned negative logarithmic distribution data can be determined as weight data. Finally, the determined weight data can be determined as the weight data group corresponding to the aforementioned target missing state array.

[0102] The ninth step is to generate structured extraction data based on the generated weighted data groups.

[0103] The above technical solution and its related content, combined with steps 101 to 107, serve as an inventive point of this disclosure, solving the problem of "high computational resource consumption." Factors leading to low data processing efficiency often include: directly extracting key information from all acquired data easily results in a large amount of data to be processed during key information extraction, leading to high data processing complexity and high computational resource consumption. Solving these factors can reduce the computational resources consumed during data processing. To achieve this effect, this disclosure firstly identifies text information data and digital information data based on the aforementioned target structured information data. This allows for the separation of different types of data for subsequent processing. Secondly, feature extraction processing is performed on the text information data to obtain the corresponding text embedding information. This yields the feature vector of the text information data. Then, feature extraction processing is performed on the digital information data to obtain the corresponding digital embedding information. This yields the feature vector of the digital information data. Finally, feature mapping processing is performed on the text embedding information and the digital embedding information to obtain text mapping information and digital mapping information. Next, the text mapping information and the number mapping information are concatenated to obtain mixed feature information. This allows for the concatenation of text mapping information and number mapping information. Then, based on the mixed feature information, missing mask data corresponding to the mixed feature information is generated. This provides clear missing information for subsequent processing. Then, based on the missing mask data, various missing state arrays are generated. Next, for each of the missing state arrays, the following steps are performed: First, the missing state array is determined as the target missing state array. Second, the missing state arrays that meet the preset processing conditions are determined as the pending missing state arrays. Then, based on the target missing state array and the pending missing state arrays, various difference data and various distribution data are generated, where each difference data corresponds to a pending missing state array, and each distribution data corresponds to a pending missing state array. Then, based on the difference data and the distribution data, weighted data groups are generated. Finally, based on the generated weighted data groups, structured extraction data is generated. Because in the process of processing target structured information data, the target structured information data can be decomposed into text information data and numerical information data, and then the text information data and numerical information data can be converted into feature vectors before processing. In this way, discrete text and numerical values ​​can be converted into fixed-length, continuous vectors, thus reducing the complexity of data processing and reducing the waste of computing resources during data processing.

[0104] In the process of adopting technical solutions to address the aforementioned technical problems, the following issues often arise:

[0105] The acquired data often contains a large amount of missing data. Directly extracting data can easily lead to a high missing rate in the extracted data, resulting in low reliability of the extracted data and wasting computing resources during data processing.

[0106] In response to the aforementioned technical problems, the following solution was adopted:

[0107] In some optional implementations of certain embodiments, the aforementioned execution entity may generate structured extraction data based on the generated weighted data groups through the following steps:

[0108] The first step is to construct a weighted adjacency matrix based on the aforementioned weighted data groups. This weighted adjacency matrix can be a matrix composed of the aforementioned weighted data groups. In practice, firstly, for each weighted data group, the executing entity can determine each weight data point in the weighted data group as a row vector. Then, the determined row vectors can be input into a matrix generation function to obtain a matrix generated from these row vectors, which serves as the weighted adjacency matrix. This matrix generation function can be any function capable of generating matrices. For example, the matrix generation function could be the `np.array()` function.

[0109] The second step involves generating a node feature matrix based on the aforementioned mixed feature information and the missing mask data. This node feature matrix can be a matrix related to the mixed feature information and the missing mask data. The node feature matrix can have 2 columns and the same number of rows as the mixed feature information and the missing mask data. For each row in the node feature matrix, the first element is the missing rate of the corresponding row in the missing mask data, and the second element is the variance of the values ​​of the elements in the corresponding row of the mixed feature information. The missing rate represents the proportion of missing data in all the data. For example, when the first row of the missing mask data is [1,0,1,0] and the first row of the mixed feature information is [2,0,2,0], the first row of the node feature matrix is ​​[0.5,1], where 0.5 is the missing rate of [1,0,1,0] and 1 is the variance of [2,0,2,0]. In practice, for each row in the missing mask data mentioned above, firstly, this row can be identified as the target row. Secondly, each element in the target row with a value of "0" can be identified as a missing element. Then, the number of missing elements can be identified as the number of missing elements. Next, the number of elements in the target row can be identified as the number of row elements. Then, the ratio of the number of missing elements to the number of row elements can be identified as the first element. Then, the rows in the mixed feature information that correspond to the target row can be identified as the target mixed row. Then, the variance of each element in the target mixed row can be identified as the second element. Then, the first element and the second element can be combined into a submatrix. Finally, the combined submatrices can be combined into a matrix as the node feature matrix.

[0110] The third step is to generate a fused feature matrix based on the aforementioned weighted adjacency matrix and node feature matrix. This fused feature matrix can be a matrix generated from the aforementioned weighted adjacency matrix and node feature matrix.

[0111] In practice, firstly, the aforementioned execution entity can input the node feature matrix into a pre-trained node feature extraction model to obtain the higher-order features corresponding to the node feature matrix as higher-order node feature information. These higher-order features can be the feature vectors corresponding to the node feature matrix. The aforementioned node feature extraction model can be a neural network model that takes the node feature matrix as input and outputs the higher-order node feature information corresponding to the node feature matrix. For example, the aforementioned node feature extraction model can be a pre-trained graph convolutional neural network. The aforementioned pre-training process can be a process of fine-tuning the graph convolutional neural network using a cross-entropy loss function and a set of node feature matrices. Then, the product of the aforementioned weight adjacency matrix and the aforementioned higher-order node feature information can be determined as the fusion feature matrix.

[0112] The fourth step involves performing feature enhancement processing on the fused feature matrix to obtain an enhanced feature matrix. This enhanced feature matrix can be the fused feature matrix after enhancement processing. In practice, the executing entity can input the fused feature matrix into an activation function to obtain the enhanced feature matrix. This activation function can be the ReLU function (Rectified Linear Unit).

[0113] Fifth, based on the enhanced feature matrix and the pre-trained classifier, generate various feature probability data. Each feature probability data point can be the missing probability corresponding to each element in each row of the enhanced feature matrix. The missing probability can be the probability corresponding to the target missing cause for each row. The target missing cause can be the missing cause with the highest probability among all missing causes. The missing cause can characterize the reason for the missing data. The missing cause can include: completely random missing, random missing, and non-random missing.

[0114] In practice, firstly, the aforementioned execution entity can input the enhanced feature matrix into a pre-trained classifier to obtain the reasons for the absence of each target. Then, the probabilities corresponding to each reason for the absence of each target can be determined as the feature probability data.

[0115] The classifier described above can be a machine learning model that takes the enhanced feature matrix as input and outputs the reasons for the missing values ​​of each target corresponding to the enhanced feature matrix.

[0116] The classifier described above can include four layers.

[0117] The first layer can be a linear feature layer. This linear feature layer takes the enhanced feature matrix as input and outputs the corresponding high-dimensional feature information. The high-dimensional feature information can be the enhanced feature matrix with a higher feature dimension. For example, the feature dimension corresponding to the high-dimensional feature information could be 8.

[0118] The second layer can be an activation function. This activation function can be a function that takes high-dimensional feature information as input and outputs the corresponding activation feature information. For example, the activation function can be the ReLU function. The activation feature information can be high-dimensional feature information processed by the ReLU function.

[0119] The third layer can be a feature mapping layer. This feature mapping layer can be a linear layer that takes activation feature information as input and outputs the corresponding low-dimensional feature information. The low-dimensional feature information can be activation feature information with an even lower feature dimension. For example, the feature dimension corresponding to the low-dimensional feature information could be 3.

[0120] The fourth layer can be the output layer. This output layer can include a normalization function and an output function. The normalization function can be a normalized exponential function that takes low-dimensional feature information as input and outputs a missing probability distribution matrix corresponding to the low-dimensional feature information. The missing probability distribution matrix can be a matrix used to characterize the missing reasons corresponding to the missing data in the enhanced feature matrix. Each row in the missing probability distribution matrix can be the missing probability distribution data of the corresponding row in the enhanced feature matrix. This missing probability distribution data can be used to characterize the missing reasons corresponding to the missing data in the enhanced feature matrix. This missing probability distribution data can include both the missing reason and the probability. The missing reason can characterize the cause of the missing data. The missing reasons can include: completely random missing, random missing, and non-random missing.

[0121] The output function described above can be an `argmax` function that takes the missing probability distribution matrix as input and outputs the corresponding target missing causes. Each of the target missing causes can be the missing cause with the highest probability in the missing probability distribution data corresponding to the missing probability distribution matrix. Each of the target missing causes can correspond to a row of data in the missing probability distribution matrix. Each of the target missing causes has a corresponding probability.

[0122] Step 6: Based on the aforementioned feature probability data, perform data repair processing on the target structured information data to obtain repaired structured information data. The repaired structured information data can be the target structured information data after data repair. The repaired structured information data includes various repair value fields. Each repair value field corresponds to a group of repair field values. Each repair value field can be a field in the repaired structured information data. Each repair field value in the group of repair field values ​​can be the field value corresponding to the repair value field.

[0123] In practice, firstly, the executing entity can determine the feature probability data with the largest corresponding value among the aforementioned feature probability data as the target feature probability data. Then, the target missing cause corresponding to the target feature probability data can be determined as the missing cause to be processed. Next, in response to determining that the missing cause to be processed is "completely random missing" or "random missing," the target structured information data can be sent to the target terminal. Then, in response to receiving the repaired target structured information data sent by the target terminal, the repaired target structured information data can be determined as the repaired structured information data. Then, in response to determining that the missing cause to be processed is "non-random missing," for each field included in the target structured information data, the fields can be divided into field groups according to a predetermined number of fields per group, serving as each partitioning value field group. Then, for each partitioning value field group, the aforementioned data processing steps can be performed using the partitioning value field group to obtain each repaired field value group corresponding to the partitioning value field group, serving as each repaired value field value group. Then, for each of the aforementioned repair value field value groups, the corresponding field can be identified as the repair value field. Finally, the identified repair value fields and the aforementioned repair value field value groups can be combined to form the repaired structured information data.

[0124] Step 7: Based on the above-mentioned structured transfer information fields, negative correlation information fields, and repair value fields included in the above-mentioned repaired structured information data, generate structured transfer information and structured negative correlation information.

[0125] In practice, the aforementioned implementing entity can determine each repair value field corresponding to each structured transfer information field within the aforementioned repair value fields as a separate transfer repair value field. Then, for each of these transfer repair value fields, firstly, the repair value field value corresponding to that field can be determined as the transfer repair value field value. Next, the transfer repair value field and its value can be combined to form value transfer-related information. Finally, the combined value transfer-related information can be combined to form structured transfer information.

[0126] In practice, the aforementioned implementing entity can determine the respective repair value fields corresponding to the aforementioned negative association information fields within each repair value field as individual negative association repair value fields. Then, for each of these negative association repair value fields, firstly, the repair value field value corresponding to that negative association repair value field can be determined as the negative association repair value field value. Next, the aforementioned negative association repair value field and its value can be combined to form value-related negative association information. Finally, the combined value-related negative association information can be combined to form structured negative association information.

[0127] Step 8: Integrate the aforementioned structured transfer information and structured negative correlation information into structured extracted data. In practice, the aforementioned structured transfer information and structured negative correlation information can be combined into structured extracted data.

[0128] The above technical solution and its related content, combined with steps 101 to 107, serve as an inventive point of this disclosure, solving the problem of "waste of computing resources." Factors leading to this waste of computing resources often include: the acquired data frequently contains a large amount of missing data; directly extracting data easily results in a high missing rate in the extracted data, leading to low reliability and thus wasting computing resources during data processing. Solving these factors can reduce the waste of computing resources. To achieve this, this disclosure first constructs a weighted adjacency matrix based on the aforementioned weighted data groups. Then, based on the aforementioned mixed feature information and the aforementioned missing mask data, a node feature matrix is ​​generated. This allows quantifying the missing state of the target structured information data by generating the node feature matrix. Next, based on the aforementioned weighted adjacency matrix and the aforementioned node feature matrix, a fusion feature matrix is ​​generated. This allows fusing the weighted adjacency matrix and the node feature matrix. Then, the fusion feature matrix undergoes feature enhancement processing to obtain an enhanced feature matrix. Finally, based on the enhanced feature matrix and a pre-trained classifier, various feature probability data are generated. Therefore, the main reasons for missing data in the target structured information data can be obtained through a classifier. Then, based on the aforementioned feature probability data, data repair processing is performed on the target structured information data to obtain repaired structured information data. This repaired structured information data includes various repair value fields, each of which corresponds to a group of repair field values. Thus, data repair processing can be performed on the target structured information data based on the reasons for missing data. Next, based on the aforementioned structured transfer information fields, negative correlation information fields, and repair value fields included in the repaired structured information data, structured transfer information and structured negative correlation information are generated. Thus, structured transfer information and structured negative correlation information can be extracted from the repaired target structured information data. Finally, the structured transfer information and structured negative correlation information are integrated into structured extracted data. This yields structured extracted data. Because missing data can be repaired using weighted adjacency matrices and node feature matrices during data processing, the reliability of the final extracted data can be improved, and the waste of computational resources during data processing can be reduced.

[0129] Step 105: Based on the obtained structured extraction data, generate target structured extraction data.

[0130] In some embodiments, the executing entity can generate target structured extraction data based on the obtained structured extraction data. The target structured extraction data can be data obtained by combining the various structured extraction data. The target structured extraction data includes target structured transfer information and target structured negative correlation information. The target structured transfer information can be data obtained by combining the various structured transfer information included in the various structured extraction data. The target structured negative correlation information can be data obtained by combining the various structured negative correlation information included in the various structured extraction data. In practice, firstly, the executing entity can combine the various structured transfer information included in the various structured extraction data into target structured transfer information. Then, it can combine the various structured negative correlation information included in the various structured extraction data into target structured negative correlation information. Finally, it can combine the target structured transfer information and the target structured negative correlation information into target structured extraction data.

[0131] Step 106: Based on the target structured extraction data, generate structured transfer information to be processed and structured negative correlation information to be processed.

[0132] In some embodiments, the aforementioned executing entity can generate structured transfer information and structured negative association information to be processed based on the aforementioned target structured extracted data. The structured transfer information to be processed can be the modified target structured transfer information. The structured transfer information to be processed includes various structured transfer fields. Each structured transfer field can be a field in the structured transfer information to be processed. Each structured transfer field corresponds to a structured transfer field value. The structured transfer field value can be the field value corresponding to the structured transfer field. Similarly, the structured negative association information to be processed can be the modified target structured negative association information. The structured negative association information to be processed includes various structured negative association fields. Each structured negative association field can be a field in the structured negative association information to be processed. Each structured negative association field corresponds to a structured negative association field value. The structured negative association field value can be the field value corresponding to the structured negative association field.

[0133] In practice, firstly, the aforementioned executing entity can display the target structured transfer information and target structured negative association information included in the target structured extracted data for technical personnel to view. Then, in response to detecting modifications made by technical personnel to the target structured transfer information and target structured negative association information included in the target structured extracted data, the modified target structured transfer information can be identified as structured transfer information to be processed. Then, the modified target structured negative association information can be identified as structured negative association information to be processed. These modification operations may include, but are not limited to, typing and deleting.

[0134] Step 107: Based on the structured transfer information to be processed and the structured negative correlation information to be processed, execute the corresponding computer resource adjustment task.

[0135] In some embodiments, the aforementioned execution entity may perform corresponding computer resource adjustment tasks based on the aforementioned structured transfer information to be processed and the aforementioned structured negative association information to be processed.

[0136] In some optional implementations of certain embodiments, the aforementioned execution entity may perform corresponding computer resource adjustment tasks based on the aforementioned structured transfer information to be processed and the aforementioned structured negative association information to be processed through the following steps:

[0137] The first step involves performing resource detection processing on the aforementioned structured transfer information and structured negative association information to obtain resource usage data. This resource usage data can be data representing the amount of memory occupied by the aforementioned structured transfer information and structured negative association information.

[0138] In practice, firstly, the executing entity can use the memory detection function to generate memory capacity data occupied by the structured transfer information to be processed as the first memory capacity data. Secondly, it can use the memory detection function to generate memory capacity data occupied by the structured negative association information to be processed as the second memory capacity data. Then, the sum of the first memory capacity data and the second memory capacity data can be determined as the resource usage data.

[0139] The second step is to determine the difference between the aforementioned resource usage data and the preset resource allocation data as the resource difference data. The aforementioned resource allocation data can be data representing the memory size pre-allocated to the aforementioned structured transfer information and the aforementioned structured negative association information to be processed.

[0140] Thirdly, in response to determining that the aforementioned resource difference data meets a preset resource difference condition, the resource quotas corresponding to the aforementioned structured transfer information and the aforementioned structured negative association information to be processed are adjusted based on the aforementioned resource adjustment data. The aforementioned resource difference condition can be that the aforementioned resource difference data is not zero. The aforementioned resource quota can represent the size of memory allocated to the aforementioned structured transfer information and the aforementioned structured negative association information to be processed. In practice, in response to determining that the aforementioned resource difference data is a positive number, the memory allocated to the aforementioned structured transfer information and the aforementioned structured negative association information to be processed can be increased using a memory adjustment function to adjust the resource quota. In response to determining that the aforementioned resource difference data is a negative number, the memory allocated to the aforementioned structured transfer information and the aforementioned structured negative association information to be processed can be decreased using the aforementioned memory adjustment function to adjust the resource quota. The aforementioned memory adjustment function can be a function capable of adjusting memory allocation. For example, the aforementioned memory adjustment function can be the malloc function.

[0141] Optionally, after step 107, the aforementioned executing entity may also perform the following steps:

[0142] The first step involves querying the structured transfer field values ​​corresponding to the structured transfer information to be processed from a pre-set value information database to obtain first query results. The value information database can be a pre-set database used to store previously generated structured transfer field values. Each first query result can be a label indicating whether a corresponding structured transfer field value exists in the value information database. For example, a first query result can be "query successful" or "query failed". In practice, for each structured transfer field value, the executing entity can query the value information database using a query function, obtaining the return value of the query function as the query return value. Then, in response to determining that the query return value is empty, "query failed" can be determined as the first query result. Finally, in response to determining that the query return value is not empty, "query successful" can be determined as the first query result. The query function can be a function capable of searching for data in a database. For example, the query function can be the `find` function.

[0143] The second step involves, for each of the aforementioned first query results, in response to determining that the first query result satisfies a preset first query condition, sending the corresponding structured transfer field and its value to the target server. The first query condition can be a query result of "query failed". The target server can be a server capable of processing the structured transfer field and its value.

[0144] Third, for each of the above first query results, in response to determining that the above first query result does not meet the above first query conditions, the following steps are performed:

[0145] The first sub-step involves controlling the corresponding alarm device to perform an alarm operation based on the structured transfer field and its value corresponding to the first query result. This alarm device can be a terminal device used by a technician. In practice, the executing entity can send the structured transfer field and its value to the alarm device to prompt the technician to review them.

[0146] The second sub-step involves, in response to receiving the transfer data cleanup information sent by the aforementioned alarm device, clearing the structured transfer fields and their values ​​corresponding to the first query result. The transfer data cleanup information can be a message prompting the executing entity to clear the structured transfer fields and their values. In practice, the executing entity can clear the structured transfer fields and their values ​​using an operation command. For example, the operation command could be `delete`.

[0147] The fourth step involves querying the value database for each structured negative association field corresponding to the structured negative association information to be processed, obtaining each second query result. Each second query result can be a label indicating whether a corresponding structured transfer field value exists in the value information database. For example, a second query result can be "query successful" or "query failed". In practice, for each structured negative association field value, the executing entity can query the value database using the query function, obtaining the return value of the query function as the query result return value. Then, in response to determining that the query result return value is empty, "query failed" can be determined as the second query result. Finally, in response to determining that the query result return value is not empty, "query successful" can be determined as the second query result.

[0148] Fifth, for each of the aforementioned second query results, in response to determining that the second query result satisfies the preset second query condition, the structured negative association field and its value corresponding to the second query result are sent to the target server. The second query condition can be that the second query result is "query failed".

[0149] Step 6: For each of the above second query results, in response to determining that the above second query result does not meet the above second query conditions, perform the following steps:

[0150] The first sub-step involves controlling the alarm device to perform corresponding alarm operations based on the structured negative association field and its value corresponding to the second query result. In practice, the executing entity can send the structured negative association field and its value to the alarm device to prompt technicians to review them.

[0151] The second sub-step involves, in response to receiving the negative correlation data cleanup information sent by the aforementioned alarm device, clearing the structured negative correlation fields and their values ​​corresponding to the second query result. The negative correlation data cleanup information can be used to prompt the executing entity to clear the structured negative correlation fields and their values. In practice, the executing entity can clear the structured negative correlation fields and their values ​​using the aforementioned operation command.

[0152] The above embodiments of this disclosure have the following beneficial effects: the computer resource adjustment method based on data source detection in some embodiments of this disclosure can improve data processing efficiency and reduce computational resource waste. Specifically, the reasons for low data processing efficiency and wasted computational resources are: when extracting key information from data, all acquired data needs to be extracted, which easily leads to low data processing efficiency. Moreover, when directly extracting key information from the acquired data, the extracted key information is prone to high repetition with previously extracted key information, resulting in wasted computational resources during extraction. Based on this, the computer resource adjustment method based on data source detection in some embodiments of this disclosure first generates various structured information data in response to receiving data acquisition information sent by the terminal. Each of the structured information data corresponds to an information data source, and each of the structured information data is data organized according to a predetermined structure. Thus, the original data to be processed can be obtained, and the acquired data can be converted into structured data through specific data acquisition tools, which can improve the efficiency of subsequent data processing. Secondly, for each information data source corresponding to the aforementioned structured information data, data detection processing is performed on the aforementioned information data source to obtain the corresponding data detection result. This allows detection of whether the data source has been updated since the last data acquisition. Then, based on the obtained data detection results, each target structured information data is determined. Therefore, only updated data can be processed. Next, for each target structured information data, data extraction processing is performed on the aforementioned target structured information data to obtain the corresponding structured extracted data, which includes structured transfer information and structured negative correlation information. Then, based on the obtained structured extracted data, target structured extracted data is generated, which includes target structured transfer information and target structured negative correlation information. This allows for the integration of the various structured extracted data. Then, based on the aforementioned target structured data extraction, unprocessed structured transfer information and unprocessed structured negative correlation information are generated. The unprocessed structured transfer information includes each structured transfer field, and the unprocessed structured negative correlation information includes each structured negative correlation field. Therefore, technical personnel can further inspect the data to remove duplicate and redundant information, improving the accuracy of the final generated data. Finally, based on the unprocessed structured transfer information and the unprocessed structured negative correlation information, corresponding computer resource adjustment tasks are executed. This allows the data to be pushed to the target server.Furthermore, after data acquisition, the data crawled from various data sources can be directly converted into structured data using a data acquisition device, and then processed, thus improving data processing efficiency. Also, before data extraction, the update time of each data source can be detected, and the results can be used to filter out data sources that have been updated. This allows for the selection of updated data from the acquired data, and only the updated data is processed, reducing the probability of duplicate data being processed repeatedly and minimizing the waste of computational resources during data processing.

[0153] Further reference Figure 2 As an implementation of the methods shown in the above figures, this disclosure provides some embodiments of a data push device, which are similar to... Figure 1 Corresponding to the method embodiments shown, the device can be specifically applied to various electronic devices.

[0154] like Figure 2As shown, a data push device 200 in some embodiments includes: a first generation unit 201, a detection unit 202, a determination unit 203, a data extraction unit 204, a second generation unit 205, a third generation unit 206, and an execution unit 207. The first generation unit 201 is configured to generate various structured information data in response to receiving data acquisition information sent by a terminal. Each structured information data corresponds to an information data source, and each structured information data is data organized according to a predetermined structure. The detection unit 202 is configured to perform data detection processing on each information data source corresponding to the structured information data to obtain a data detection result corresponding to the information data source. The determination unit 203 is configured to determine various target structured information data based on the obtained data detection results. The data extraction unit 204 is configured to perform data extraction on each target structured information data. The data is processed to obtain structured extracted data corresponding to the aforementioned target structured information data, wherein the structured extracted data includes structured transfer information and structured negative correlation information; the first generation unit 205 is configured to generate target structured extracted data based on the obtained structured extracted data, wherein the target structured extracted data includes target structured transfer information and target structured negative correlation information; the second generation unit 206 is configured to generate unprocessed structured transfer information and unprocessed structured negative correlation information based on the aforementioned target structured extracted data, wherein the unprocessed structured transfer information includes various structured transfer fields, and the unprocessed structured negative correlation information includes various structured negative correlation fields; the execution unit 207 is configured to execute corresponding computer resource adjustment tasks based on the unprocessed structured transfer information and the unprocessed structured negative correlation information.

[0155] It is understandable that the units described in the data push device 200 are related to the reference. Figure 1 The steps in the described method correspond to each other. Therefore, the operations, features, and beneficial effects described above for the method also apply to the data push device 200 and the units contained therein, and will not be repeated here.

[0156] The following is for reference. Figure 3 It shows a schematic diagram of the structure of an electronic device (such as a computing device) 300 suitable for implementing some embodiments of the present disclosure. Figure 3 The electronic device shown is merely an example and should not be construed as limiting the functionality and scope of the embodiments of this disclosure.

[0157] like Figure 3As shown, the electronic device 300 may include a processing unit (e.g., a central processing unit, a graphics processing unit, etc.) 301, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 302 or a program loaded from a storage device 308 into a random access memory (RAM) 303. The RAM 303 also stores various programs and data required for the operation of the electronic device 300. The processing unit 301, ROM 302, and RAM 303 are interconnected via a bus 304. An input / output (I / O) interface 305 is also connected to the bus 304.

[0158] Typically, the following devices can be connected to I / O interface 305: input devices 306 including, for example, touchscreens, touchpads, keyboards, mice, cameras, microphones, accelerometers, gyroscopes, etc.; output devices 307 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; storage devices 308 including, for example, magnetic tapes, hard disks, etc.; and communication devices 309. Communication device 309 allows electronic device 300 to communicate wirelessly or wiredly with other devices to exchange data. Although Figure 3 An electronic device 300 with various devices is shown; however, it should be understood that it is not required to implement or possess all of the devices shown. More or fewer devices may be implemented or possessed alternatively. Figure 3 Each box shown can represent a device or multiple devices as needed.

[0159] In particular, according to some embodiments of this disclosure, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, some embodiments of this disclosure include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via communication device 309, or installed from storage device 308, or installed from ROM 302. When the computer program is executed by processing device 301, it performs the functions defined in the methods of some embodiments of this disclosure.

[0160] It should be noted that, in some embodiments of this disclosure, the computer-readable medium may be a computer-readable signal medium or a computer-readable storage medium, or any combination thereof. A computer-readable storage medium may be, for example,—but not limited to—an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of a computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In some embodiments of this disclosure, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. In some embodiments of this disclosure, a computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium can be any computer-readable medium other than a computer-readable storage medium, which can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium can be transmitted using any suitable medium, including but not limited to: wires, optical fibers, RF (radio frequency), etc., or any suitable combination thereof.

[0161] In some implementations, clients and servers can communicate using any currently known or future-developed network protocol such as HTTP (Hypertext Transfer Protocol) and can interconnect with digital data communication (e.g., communication networks) of any form or medium. Examples of communication networks include local area networks (“LANs”), wide area networks (“WANs”), the Internet (e.g., the Internet of Things), and end-to-end networks (e.g., ad hoc end-to-end networks), as well as any currently known or future-developed networks.

[0162] The aforementioned computer-readable medium may be included in the aforementioned electronic device; or it may exist independently and not assembled into the electronic device. The aforementioned computer-readable medium carries one or more programs, which, when executed by the electronic device, cause the electronic device to: generate various structured information data in response to receiving data acquisition information sent by a terminal, wherein each of the aforementioned structured information data corresponds to an information data source, and each of the aforementioned structured information data is data organized according to a predetermined structure; for each of the information data sources corresponding to the aforementioned structured information data, perform data detection processing on the information data source to obtain a data detection result corresponding to the information data source; based on the obtained data detection results, determine various target structured information data; for each of the aforementioned target structured information data, The aforementioned target structured information data is processed to extract structured data, which includes structured transfer information and structured negative correlation information. Based on the obtained structured extract data, target structured extract data is generated, which includes target structured transfer information and target structured negative correlation information. Based on the target structured extract data, unprocessed structured transfer information and unprocessed structured negative correlation information are generated, whereby the unprocessed structured transfer information includes various structured transfer fields, and the unprocessed structured negative correlation information includes various structured negative correlation fields. Based on the unprocessed structured transfer information and the unprocessed structured negative correlation information, corresponding computer resource adjustment tasks are executed.

[0163] Computer program code for performing operations of some embodiments of this disclosure can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0164] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0165] The units described in some embodiments of this disclosure can be implemented in software or hardware. The described units can also be housed in a processor; for example, a processor may be described as including a first generation unit, a detection unit, a determination unit, a data extraction unit, a second generation unit, a third generation unit, and an execution unit. The names of these units do not necessarily limit the unit itself; for example, the execution unit may also be described as "a unit that performs corresponding computer resource adjustment tasks."

[0166] The functions described above in this document can be performed, at least in part, by one or more hardware logic components. For example, exemplary types of hardware logic components that can be used, without limitation, include: Field Programmable Gate Arrays (FPGAs), Application-Specific Integrated Circuits (ASICs), Application Standard Products (ASSPs), System-on-Chip (SoCs), Complex Programmable Logic Devices (CPLDs), and so on.

[0167] The above description is merely a selection of preferred embodiments of this disclosure and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of the invention involved in the embodiments of this disclosure is not limited to technical solutions formed by specific combinations of the above-described technical features, but should also cover other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the above-described inventive concept. For example, technical solutions formed by substituting the above-described features with (but not limited to) technical features with similar functions disclosed in the embodiments of this disclosure.

Claims

1. A method for adjusting computer resources based on data source detection, comprising: In response to receiving data acquisition information sent by the terminal, various structured information data are generated, wherein each structured information data corresponds to an information data source, and each structured information data is data organized according to a predetermined structure; For each information data source corresponding to each structured information data, data detection processing is performed on the information data source to obtain the data detection result corresponding to the information data source; Based on the obtained data detection results, the structured information data of each target is determined; For each target structured information data in the aforementioned target structured information data, data extraction processing is performed on the target structured information data to obtain the corresponding structured extracted data, wherein the structured extracted data includes structured transfer information and structured negative correlation information, including: Based on the target structured information data, determine the text information data and the digital information data; The text information data is subjected to feature extraction processing to obtain the text embedding information corresponding to the text information data; The digital information data is subjected to feature extraction processing to obtain the digital embedded information corresponding to the digital information data; The text embedding information and the number embedding information are subjected to feature mapping processing to obtain text mapping information and number mapping information; The text mapping information and the number mapping information are subjected to feature concatenation processing to obtain mixed feature information; Based on the hybrid feature information, generate missing mask data corresponding to the hybrid feature information; Based on the missing mask data, generate arrays of various missing states; For each of the missing state arrays, perform the following steps: The missing state array is determined as the target missing state array; Each missing state array that meets the preset processing conditions is determined as a missing state array to be processed. Based on the target missing state array and the various missing state arrays to be processed, generate various difference data and various distribution data; Based on the aforementioned differential data and distribution data, a weighted data group is generated; Based on the generated weighted data groups, structured extraction data is generated; Based on the obtained structured extraction data, target structured extraction data is generated, wherein the target structured extraction data includes target structured transfer information and target structured negative correlation information; Based on the target structured extracted data, structured transfer information to be processed and structured negative association information to be processed are generated. The structured transfer information to be processed includes each structured transfer field, and the structured negative association information to be processed includes each structured negative association field. Based on the structured transfer information to be processed and the structured negative correlation information to be processed, the corresponding computer resource adjustment task is executed.

2. The method according to claim 1, wherein, The step of performing corresponding computer resource adjustment tasks based on the structured transfer information to be processed and the structured negative correlation information to be processed includes: Resource detection processing is performed on the structured transfer information and the structured negative correlation information to be processed to obtain resource consumption data; The difference between the resource consumption data and the preset resource allocation data is determined as the resource difference data; In response to determining that the resource difference data meets the preset resource difference conditions, the resource quota corresponding to the structured transfer information to be processed and the structured negative association information to be processed is adjusted based on the resource difference data.

3. The method according to claim 1, wherein, Based on the obtained data detection results, the structured information data of each target is determined, including: Based on the obtained data detection results, each data detection result that meets the preset data detection conditions is determined as each target detection result. Based on the detection results of each target, the structured information data corresponding to each target detection result is determined as the structured information data of each target.

4. The method according to claim 1, wherein, Each of the various information data sources has a corresponding data update time; The process of performing data detection processing on the information data source to obtain the data detection result corresponding to the information data source includes: Based on the information data source, the pre-stored update times of various historical data, and the data update time, an update comparison result is generated; Based on the structured information data corresponding to the information data source, update flag data corresponding to the information data source is generated; Based on the preset historical update flag data and the update flag data, generate flag data comparison results; Based on the updated comparison results and the flag data comparison results, data detection results are generated.

5. The method according to claim 1, wherein, Each target structured information data includes various fields, and each field corresponds to a group of field values. The process of extracting data from the target structured information data to obtain structured extracted data corresponding to the target structured information data includes: The target structured information data is divided into various fields to obtain various field groups, wherein each field in each field group corresponds to a field value group. For each of the partition field groups, perform the following data processing steps: Perform missing value detection processing on each field value group corresponding to the divided field group to obtain each missing value detection result group corresponding to each field value group; For each missing value detection result in each missing value detection result group, in response to determining that the missing value detection result meets the preset missing condition, the following steps are performed: The field value corresponding to the missing value detection result is determined as the target field value; In response to determining that the field corresponding to the missing value detection result meets the preset field condition, the preset missing value is determined as the target field value in the partitioned field group, so as to update the partitioned field group; In response to determining that the missing value detection result does not meet the field condition, the following steps are performed: Based on the missing value detection results and pre-stored historical target structured information data, determine the historical field values ​​corresponding to the missing value detection results; The historical field value is determined as the target field value in the partition field group in order to update the partition field group; Each partition field in the updated partition field group that meets the preset numeric field conditions is determined as a field to be detected, wherein each field to be detected corresponds to a field value group to be detected. Based on a preset field unit information table, data standardization processing is performed on each field value group corresponding to each field to be detected to obtain each standardized field value group. Based on a preset set of field range information, abnormal field detection processing is performed on each standardized field value set to obtain an abnormal field detection result set corresponding to each standardized field value set. Based on the detection result groups of each abnormal field, data repair processing is performed on each standardized field value group to obtain each repaired field value group, wherein each repaired field value group corresponds to a repaired field. The repair fields corresponding to the respective repair field value groups are combined into a repair field group; Based on the preset structured transfer information fields, the preset negative association information fields, and the obtained repair field groups, structured transfer information and structured negative association information are generated. The structured transfer information and the structured negative correlation information are combined to form structured extracted data corresponding to the target structured information data.

6. The method according to claim 5, wherein, The process of performing data repair on each standardized field value group based on the detection result group of each abnormal field to obtain each repaired field value group includes: For each anomaly detection result group in the aforementioned anomaly field detection result groups, perform the following steps: For each anomaly detection result in the anomaly detection result group, in response to determining that the anomaly detection result meets the preset anomaly field conditions, the following steps are performed: The standardized field value corresponding to the anomaly field detection result is determined as the target standardized field value; The field range information corresponding to the abnormal field detection result in the field range information group is determined as the target field range information; Based on the target field range information, determine the average value of the field corresponding to the target field range information; The average value of the field is determined as the target standardized field value, and the target standardized field value is updated accordingly. The updated target standardized field values ​​are determined as the target standardized field value group corresponding to the anomaly field detection result group; The target standardized field value group is identified as the repair field value group.

7. The method according to claim 1, wherein, Each structured transfer field in the structured transfer field has a corresponding structured transfer field value, and each structured negative association field in the structured negative association field has a corresponding structured negative association field value. The method also includes: Based on the structured transfer field values ​​corresponding to the structured transfer information to be processed, query the structured transfer field values ​​from the preset value information database to obtain the first query results. For each of the first query results, in response to determining that the first query result satisfies the preset first query condition, the structured transfer field and the structured transfer field value corresponding to the first query result are sent to the target server. For each of the first query results, in response to determining that the first query result does not meet the first query condition, the following steps are performed: Based on the structured transfer field and structured transfer field value corresponding to the first query result, control the corresponding alarm device to perform the corresponding alarm operation; In response to receiving the transfer data cleanup information sent by the alarm device, the structured transfer field and structured transfer field value corresponding to the first query result are cleared. Based on the values ​​of each structured negative association field corresponding to the structured negative association information to be processed, query the values ​​of each structured negative association field from the value information database to obtain each second query result; For each of the second query results, in response to determining that the second query result satisfies the preset second query condition, the structured negative association field and the structured negative association field value corresponding to the second query result are sent to the target server; For each of the second query results, in response to determining that the second query result does not satisfy the second query condition, the following steps are performed: Based on the structured negative association field and structured negative association field value corresponding to the second query result, the alarm device is controlled to perform the corresponding alarm operation; In response to receiving the negative correlation data cleanup information sent by the alarm device, the structured negative correlation field and the structured negative correlation field value corresponding to the second query result are cleared.

8. A computer resource adjustment device based on data source detection, comprising: The first generation unit is configured to generate various structured information data in response to receiving data acquisition information sent by the terminal, wherein each structured information data corresponds to an information data source, and each structured information data is data organized according to a predetermined structure. The detection unit is configured to perform data detection processing on each information data source in each information data source corresponding to each structured information data, and obtain the data detection result corresponding to the information data source; The determining unit is configured to determine the structured information data of each target based on the obtained detection results of each data. The data extraction unit is configured to perform data extraction processing on each target structured information data in the respective target structured information data to obtain structured extracted data corresponding to the target structured information data. The structured extracted data includes structured transition information and structured negative correlation information, including: determining text information data and digital information data based on the target structured information data; performing feature extraction processing on the text information data to obtain text embedding information corresponding to the text information data; performing feature extraction processing on the digital information data to obtain digital embedding information corresponding to the digital information data; performing feature mapping processing on the text embedding information and the digital embedding information to obtain text mapping information and digital mapping information; and performing feature mapping on the text mapping information. The radiometric information and the digital mapping information are subjected to feature concatenation processing to obtain mixed feature information; based on the mixed feature information, missing mask data corresponding to the mixed feature information is generated; based on the missing mask data, various missing state arrays are generated; for each missing state array in the various missing state arrays, the following steps are performed: the missing state array is determined as the target missing state array; each missing state array in the various missing state arrays that meets the preset processing conditions is determined as a missing state array to be processed; based on the target missing state array and the various missing state arrays to be processed, various difference data and various distribution data are generated; based on the various difference data and the various distribution data, weighted data groups are generated; based on the generated weighted data groups, structured extraction data is generated; The second generation unit is configured to generate target structured extraction data based on the obtained structured extraction data, wherein the target structured extraction data includes target structured transfer information and target structured negative correlation information. The third generation unit is configured to generate structured transfer information to be processed and structured negative association information to be processed based on the target structured extracted data. The structured transfer information to be processed includes various structured transfer fields, and the structured negative association information to be processed includes various structured negative association fields. The execution unit is configured to perform corresponding computer resource adjustment tasks based on the structured transfer information to be processed and the structured negative association information to be processed.

9. An electronic device, comprising: One or more processors; A storage device on which one or more programs are stored; When the one or more programs are executed by the one or more processors, the one or more processors implement the method as described in any one of claims 1 to 7.

10. A computer-readable medium having a computer program stored thereon, wherein, When the program is executed by the processor, it implements the method as described in any one of claims 1 to 7.

Images