Data storage information security processing system and method based on cloud computing

Abstract

The invention relates to the field of information security storage, in particular to a data storage information security processing system and method based on cloud computing, and the method comprises the steps: reading target analysis data, setting a search word, retrieving homologous data in a storage database, setting a data structure according to a data scene, and storing the data in a database; determining an inevitable structure by combining the structure occurrence frequency in the homologous data so as to judge a structure complete signal; the structural data segment is divided into public data and novel data, risk coefficients are analyzed according to the occurrence frequency of the public data, the number of source abnormity times and historical output data of a novel data source, then a comprehensive risk value is determined to distinguish storage and safety abnormal signals, data are displayed when the signals are abnormal, and acousto-optic reminding is generated; according to the method, the limitation of single data processing is solved, the integrity of a data structure is guaranteed, the risk assessment accuracy is improved, the exception response efficiency is optimized, and the safety and reliability of cloud computing data storage are practically guaranteed.

Description

Technical Field

[0001] This invention relates to the field of information security storage technology, and in particular to a data storage information security processing system and method based on cloud computing. Background Technology

[0002] With the rapid development of cloud computing technology, the amount of data generated is growing exponentially. Enterprises and individuals have an increasingly urgent need for elastic expansion, low cost, and high availability of data storage. Cloud computing data storage, with its advantages of resource pooling and on-demand allocation, has become the mainstream data storage model.

[0003] However, existing risk assessment methods often use a uniform standard to assess the data as a whole, without distinguishing between commonly used data and data that appears for the first time or very rarely in the same source data. This leads to large deviations in risk assessment results and is prone to misjudging safety or missing risks. Summary of the Invention

[0004] The purpose of this invention is to solve the problems in the background art by proposing a data storage information security processing system and method based on cloud computing.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: A cloud computing-based data storage and information security processing system includes: The data reading module is used to read the data to be stored and mark it as target analysis data; The data retrieval module is used to convert target analysis data into text form, set search terms, and perform big data retrieval and processing in the storage database based on the search terms to identify data from the same source. The structural analysis module is used to set the data structure for the target analysis data, obtain the same source data, determine the necessary structure in the data structure based on the frequency of the data structure in the same source data, and then determine the structural integrity signal of the target analysis data based on the necessary structure. The risk identification module is used to acquire target analysis data corresponding to structurally complete signals. Based on the data structure of the target analysis data, the data information corresponding to each data structure is marked as a structural data segment. Then, based on the same source data, the structural data segments are divided into common data and new data. Public data and new data are analyzed separately to determine the risk coefficients of public data and new data. Finally, based on the risk coefficients of public data and new data, the comprehensive risk value of the target analysis data is determined. Based on the comprehensive risk value, storage signals and security anomaly signals are then determined. The information storage module is used to store the target analysis data corresponding to the stored signal; The anomaly display module is used to display the target analysis data corresponding to the safety anomaly signal on the display terminal, and at the same time generate audio and visual reminder information.

[0006] Preferred methods for determining source data include: The target analysis data is converted into text using a multimodal Transformer model. The text data is then split into words, and the resulting content words are marked as keywords. These keywords are then used as search terms, and a large-scale data search is performed in the storage database to obtain the search data. The retrieved data is sequentially compared with the target analysis data to calculate the similarity value. The information similarity value is then compared with the information similarity threshold A1. If the information similarity value is greater than or equal to the information similarity threshold A1, the corresponding retrieved data is marked as source data. Conversely, if the information similarity value is less than the information similarity threshold A1, the corresponding retrieved data is marked as irrelevant information.

[0007] Preferred methods for determining structurally intact signals include: Acquire all data from the same source, analyze the data scenario according to the target data, and identify the data structure in each piece of data from the same source. Here, the data scenario refers to the specific business background of data generation, flow, and application, and the data structure is the organization form of information in the data scenario. All data structures are sequentially labeled as target structures. Source data is obtained, and the existence of target structures in source data is identified. If target structures exist in source data, a presence label is set for this source data. All source data is traversed, and the number of source data containing the presence label is counted and marked as the structure occurrence frequency TC. The structure occurrence frequency FC of the target structure is obtained by using the formula TC÷ZH=FC, where ZH represents the total number of single source data. The frequency of occurrence of the structure FC is compared with the frequency threshold X1. If FC < X1, the target structure is marked as an accidental structure; otherwise, if FC ≥ X1, the target structure is marked as a necessary structure. Extract all necessary structures from the data structure, then obtain the target analysis data, identify the data structure in the target analysis data. If all necessary structures exist in the data structure of the target analysis data, a complete structure signal is generated. Conversely, if one or more necessary structures do not appear in the data structure of the target analysis data, a missing structure signal is generated.

[0008] Preferably, when a structural missing signal is generated, the corresponding target analysis data is transmitted to the anomaly display module. The anomaly display module displays the target analysis data corresponding to the structural anomaly signal on the display terminal and generates an audio-visual reminder message. The corresponding data management personnel then re-confirm the safety and structural integrity of the target analysis data.

[0009] Preferably, the risk identification module includes a data processing unit, a common analysis unit, a difference analysis unit, and a comprehensive processing unit; The data processing unit is used to acquire target analysis data corresponding to the structurally complete signal. Based on the data structure of the target analysis data, the data information corresponding to each data structure is marked as a structural data segment. Then, based on the same source data, the structural data segments are divided into common data and new data. The public analysis unit is used to analyze public data based on the frequency of public data occurrence in the same source data and the number of data anomalies in the data source location of the public data, and to determine the risk coefficient of the public data. The difference analysis unit is used to identify the data source location of the new data, extract historical output data from this data source location, analyze the historical output data, and determine the risk coefficient of the new data. The integrated processing unit is used to determine the comprehensive risk value of the target analysis data based on the risk coefficients of public data and new data, and then determine the storage signal and security anomaly signal based on the comprehensive risk value.

[0010] Preferred methods for determining public data and novel data include: Obtain target analysis data and source data, divide the target analysis data according to data structure, and mark the data information corresponding to each data structure as a structure data segment; The structured data segment is searched in the same source information. If the structured data segment exists in the same source information, it is marked as public data. Otherwise, if the structured data segment does not exist in the same source information, it is marked as new data.

[0011] Preferably, the methods for determining the risk coefficient of novel data include: Select any common data b, identify the number of times common data b appears in the same source data, and mark it as the number of segments. Divide the number of segments of common data b by the total number of the same source data, and mark the result as the number of segments frequency FDb. Then identify the data source location of public data b, and set a time interval according to the periodic time, with the current time as the time node. The current time is the end time of the time interval, and the periodic time is the length of this time interval. Within a time interval, count the number of data anomalies in the data source locations of public data b, and divide this number of data anomalies by the number of data produced in this data source location to obtain the location anomaly frequency FWb. Using calculation formula Obtain the anomaly coefficient Yb of the public data b, where, This represents the error coefficient indicating the location of the data source. All public data are processed according to the above method to obtain the error coefficient of each public data. In the target analysis data, the error coefficients of all public data are obtained, and the mean of the error coefficients is calculated. The result of the mean calculation is marked as the risk coefficient Gd of the public data.

[0012] Preferably, the methods for determining the risk coefficient of novel data include: Based on the data source location of the new data, historical output data from this data source location is extracted. Then, the new data and historical output data are processed for similarity to obtain data similarity value. Historical output data with data similarity value greater than or equal to the data similarity threshold X2 are selected and marked as similar control data. Obtain all similar control data and perform mean processing on the similar control data. Mark the obtained mean result as the control characteristic value. Then, subtract the control characteristic value from the new data to obtain the offset value. Take the absolute value of the offset value and mark the absolute value as the offset representative value. The offset representative value of all new data is calculated according to the above method. Then, based on all the new data in the target analysis data, the offset representative value of the new data is averaged and the resulting value is marked as the risk coefficient Gx of the new data. When the data source of the new data has not generated any data before, the risk coefficient of the new data is directly set to m, that is, Gx=m.

[0013] Preferably, the method for determining the stored signal and the security anomaly signal includes: Obtain the risk coefficient Gd of the common data and the risk coefficient Gx of the novel data in the target analysis data, and then use the calculation formula. The comprehensive risk value Zf of the target analysis data is obtained, where, and These are the proportionality coefficients, and ; The comprehensive risk value Zf of the target analysis data is compared with the risk threshold K1. If Zf ≤ K1, a storage signal is generated and the target analysis data corresponding to the storage signal is transmitted to the information storage unit. Conversely, if Zf > K1, a safety anomaly signal is generated and transmitted to the anomaly display module.

[0014] A cloud computing-based method for secure data storage processing, comprising the following steps: Step 1: Read the data to be stored and mark it as target analysis data. Convert the target analysis data into text form and set search terms. Based on the search terms, perform big data retrieval and processing in the storage database to identify data from the same source. Step 2: Based on the data scenario of the target analysis data, set the data structure, obtain the same source data, determine the necessary structure in the data structure based on the frequency of the data structure in the same source data, and then determine the structural integrity signal of the target analysis data based on the necessary structure. Step 3: Obtain the target analysis data corresponding to the structurally complete signal. Based on the data structure of the target analysis data, mark the data information corresponding to each data structure as a structural data segment. Then, based on the same source data, divide the structural data segments into common data and new data. Step 4: Based on the frequency of public data occurrences in the same source data and the number of data anomalies in the data source location of the public data, analyze the public data to determine the risk coefficient of the public data. Then, identify the data source location of the new data, extract the historical output data from this data source location, analyze the historical output data to determine the risk coefficient of the new data, and finally, based on the risk coefficients of the public data and the new data, determine the comprehensive risk value of the target analysis data. Based on the comprehensive risk value, determine the storage signal and security anomaly signal. Step 5: When a stored signal is detected, the target analysis data corresponding to the stored signal is stored in the data storage. When a security anomaly signal is detected, the corresponding target analysis data is displayed on the display terminal, and an audio-visual reminder message is generated at the same time.

[0015] Compared with existing technologies, the advantages of this invention are: This invention reads target analysis data and sets search terms to retrieve source data from the storage database. Then, it sets the data structure according to the data scenario and determines the necessary structure by combining the frequency of structure occurrence in the source data to judge the structural integrity signal. This enables accurate identification of structurally incomplete or abnormal target data, preventing such data from entering the storage stage and significantly improving the structural reliability and subsequent usability of stored data. Next, the structured data segments are divided into common and novel data. Risk coefficients are analyzed based on the frequency of occurrence and source anomalies of common data, and the historical output data of novel data sources, respectively. A comprehensive risk value is then determined to distinguish between storage and security anomaly signals. When an anomaly occurs, the data is displayed and an audio-visual alert is generated. By combining the risk coefficients of the two types of data to obtain the comprehensive risk value, this invention effectively solves the problem of existing technologies having a single and incomplete risk assessment standard, significantly improving the accuracy of risk assessment. This solution overcomes the limitations of existing technologies that only process data, ensures data structure integrity, improves the accuracy of risk assessment, optimizes anomaly response efficiency, and effectively ensures the security and reliability of cloud computing data storage. Attached Figure Description

[0016] Figure 1 This is a schematic diagram of the system structure of the present invention; Figure 2 This is a schematic diagram of the method flow of the present invention. Detailed Implementation

[0017] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0018] Reference Figure 1 and Figure 2 A cloud computing-based data storage and information security processing system includes a data reading module, a data retrieval module, a structure analysis module, a risk identification module, an information storage module, and an anomaly display module. The data reading module is used to read the data information to be stored and mark the data information to be stored as target analysis data. Then the data reading module transmits the target analysis data to the data retrieval module. The data retrieval module receives target analysis data, converts it into text, then splits the text into words, marks the split content words as keywords, uses the keywords as search terms, and performs big data retrieval in the storage database to obtain the retrieved data. It should be further explained that the target analysis data is converted into text form using a multimodal Transformer model. That is, a unified Transformer architecture is used to process inputs from different modalities, cross-modal information is fused through a self-attention mechanism, and then a unified text is generated. The target analysis data in text form is split into words using a word segmentation algorithm based on deep learning, such as a recurrent neural network. A temporal model is used to capture contextual semantics and output word segmentation labels. The content words here are words in Chinese vocabulary that have actual lexical meaning and can independently function as sentence components, such as subject, predicate, and object. The obtained search data is sequentially compared with the target analysis data to calculate the similarity value. The information similarity value is then compared with the information similarity threshold A1. If the information similarity value is greater than or equal to the information similarity threshold A1, the corresponding search data is marked as source data. Conversely, if the information similarity value is less than the information similarity threshold A1, the corresponding search data is marked as irrelevant information. The specific value of the information similarity threshold A1 is obtained by those skilled in the art through big data calculation. Then, a one-way communication connection is established between the data retrieval module and the structural analysis module, and the target analysis data and the same source data are transmitted to the structural analysis module. The structural analysis module is used to analyze the data structure of the target analysis data and determine its structural integrity. Specific methods for determining the structural integrity of the target analysis data include: S1: Obtain all source data and identify the data structure in each source data according to the data scenario of the target analysis data. The data scenario refers to the specific business background of data generation, flow and application, and the data structure is the organization form of information in the data scenario. For example, when the data scenario of the target analysis data is business data, the data structure of the business data includes order ID, user ID, product name, amount and order status, etc. Choose any data structure and take this data structure as an example. Mark this data structure as the target structure, obtain the same source data, identify whether the target structure exists in the same source data. If the target structure exists in the same source data, set the existence label for this same source data. Traverse all the same source data, then count the number of same source data containing the existence label and mark it as the structure occurrence frequency TC. Then use the formula TC÷ZH=FC to get the structure occurrence frequency FC of the target structure, where ZH represents the total number of single same source data. The frequency FC of the structure is compared with the frequency threshold X1. If FC < X1, the target structure is marked as an accidental structure. Conversely, if FC ≥ X1, the target structure is marked as a necessary structure. The specific value of the frequency threshold X1 is obtained by those skilled in the art based on big data calculations. All data structures are sequentially labeled as target structures and processed according to the above method. At this point, all data structures are divided into accidental structures and inevitable structures. S2: Extract all necessary structures from the data structure, then obtain the target analysis data, identify the data structure in the target analysis data. If all necessary structures exist in the data structure of the target analysis data, generate a complete structure signal. Otherwise, if one or more necessary structures do not appear in the data structure of the target analysis data, generate a missing structure signal. Then, a one-way communication connection is established between the structural analysis module, the risk identification module, and the anomaly display module. When a structural integrity signal is generated, the corresponding target analysis data is transmitted to the risk identification module. If a structural missing signal is generated, the corresponding target analysis data is transmitted to the anomaly display module. The risk identification module is used to receive target analysis data corresponding to structurally intact signals and perform risk analysis on the target analysis data. Furthermore, the risk identification module includes a data processing unit, a common analysis unit, a difference analysis unit, and a comprehensive processing unit. The data processing unit is used to acquire target analysis data and source data, divide the target analysis data according to the data structure, and mark the data information corresponding to each data structure as a structure data segment; The structured data segment is searched in the same source information. If the structured data segment exists in the same source information, it is marked as public data. Otherwise, if the structured data segment does not exist in the same source information, it is marked as new data. The data processing unit then transmits the common data from the target analysis data to the common analysis unit and transmits the novel data from the target analysis data to the difference analysis unit. The common analysis unit is used to receive common data from the target analysis data, analyze the common data, and determine the risk coefficient of the common data. Specifically, the methods for determining the risk coefficient of the common data include: Take any common data b as an example, identify the number of times common data b appears in the same source data, and mark it as the number of segments. Divide the number of segments of common data b by the total number of the same source data, and mark the result as the number of segments frequency FDb. Then identify the data source location of public data b, and set a time interval according to the periodic time, with the current time as the time node. The current time is the end time of the time interval, and the periodic time is the length of this time interval. Furthermore, the periodic time is a threshold, and the specific length of the time is set by those skilled in the art based on big data experience. Within a time interval, count the number of data anomalies in the data source locations of public data b, and divide this number of data anomalies by the number of data produced in this data source location to obtain the location anomaly frequency FWb. Using calculation formula Obtain the anomaly coefficient Yb of the public data b, where, This represents the error coefficient for the data source location. The specific values ​​were obtained by those skilled in the art through big data calculations; All public data are processed according to the above method to obtain the error coefficient of each public data. In the target analysis data, the error coefficients of all public data are obtained, and the mean of the error coefficients is calculated. The mean calculation result is marked as the risk coefficient Gd of the public data. Then, a one-way communication connection is established between the common analysis unit and the integrated processing unit, and the risk coefficient of the common data is transmitted to the integrated processing unit; The difference analysis unit is used to receive novel data from the target analysis data, analyze the novel data, and determine the risk coefficient of the novel data. Specifically, the methods for determining the risk coefficient of novel data include: Based on the data source location of the new data, historical output data from this data source location is extracted. Then, the new data and historical output data are processed for similarity to obtain a data similarity value. Historical output data with a data similarity value greater than or equal to the data similarity threshold X2 are selected and marked as similar control data. In this embodiment, the similarity processing algorithm is a cosine similarity algorithm, and the specific value of the data similarity threshold X2 is obtained by those skilled in the art based on big data calculations. Obtain all similar control data and perform mean processing on the similar control data. Mark the obtained mean result as the control characteristic value. Then, subtract the control characteristic value from the new data to obtain the offset value. Take the absolute value of the offset value and mark the absolute value as the offset representative value. Calculate the offset representative value of all new data according to the above method, and then, based on all new data in the target analysis data, perform mean processing on the offset representative value of the new data, and mark the obtained value as the risk coefficient Gx of the new data; It should be further explained that when the source of the new data has not generated any data before, the risk coefficient of the new data is directly set to m, that is, Gx=m. Furthermore, m is the maximum risk threshold, and the specific value of m is obtained by those skilled in the art based on big data calculations. Then, a one-way communication connection is established between the difference analysis unit and the integrated processing unit, and the risk coefficient of the new data is transmitted to the integrated processing unit. The integrated processing unit receives the risk coefficient Gd of common data and the risk coefficient Gx of novel data from the target analysis data, and then uses the calculation formula... The comprehensive risk value Zf of the target analysis data is obtained, where, and These are the proportionality coefficients, and Furthermore, and The specific values ​​were obtained by those skilled in the art based on big data calculations. At the same time, the larger the comprehensive risk value Zf is, the greater the risk of data anomalies in the target analysis data. Conversely, the smaller the comprehensive risk value Zf is, the smaller the risk of data anomalies in the target analysis data. The comprehensive risk value Zf of the target analysis data is compared with the risk threshold K1. If Zf≤K1, a storage signal is generated and the target analysis data corresponding to the storage signal is transmitted to the information storage unit. Conversely, if Zf>K1, a safety anomaly signal is generated and transmitted to the anomaly display module. The specific value of the risk threshold K1 is set by those skilled in the art based on big data experience. The information storage module is used to detect the stored signal and store the target analysis data corresponding to the stored signal. The anomaly display module is used to detect safety anomaly signals and structural anomaly signals, and displays the target analysis data corresponding to the safety anomaly signals and structural anomaly signals on the display terminal. At the same time, it generates audio and visual reminder information, which allows the corresponding data management personnel to reconfirm the safety and structural integrity of the target analysis data.

[0019] A cloud computing-based method for secure data storage processing, comprising the following steps: Step 1: Read the data to be stored and mark it as target analysis data. Convert the target analysis data into text form and set search terms. Based on the search terms, perform big data retrieval and processing in the storage database to identify data from the same source. Step 2: Based on the data scenario of the target analysis data, set the data structure, obtain the same source data, determine the necessary structure in the data structure based on the frequency of the data structure in the same source data, and then determine the structural integrity signal of the target analysis data based on the necessary structure. Step 3: Obtain the target analysis data corresponding to the structurally complete signal. Based on the data structure of the target analysis data, mark the data information corresponding to each data structure as a structural data segment. Then, based on the same source data, divide the structural data segments into common data and new data. Step 4: Based on the frequency of public data occurrences in the same source data and the number of data anomalies in the data source location of the public data, analyze the public data to determine the risk coefficient of the public data. Then, identify the data source location of the new data, extract the historical output data from this data source location, analyze the historical output data to determine the risk coefficient of the new data, and finally, based on the risk coefficients of the public data and the new data, determine the comprehensive risk value of the target analysis data. Based on the comprehensive risk value, determine the storage signal and security anomaly signal. Step 5: When a stored signal is detected, the target analysis data corresponding to the stored signal is stored in the data storage. When a security anomaly signal is detected, the corresponding target analysis data is displayed on the display terminal, and an audio-visual reminder message is generated at the same time.

[0020] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A cloud computing-based data storage and information security processing system, characterized in that, include: The data reading module is used to read the data to be stored and mark it as target analysis data; The data retrieval module is used to convert target analysis data into text form, set search terms, and perform big data retrieval and processing in the storage database based on the search terms to identify data from the same source. The structural analysis module is used to set the data structure for the target analysis data, obtain the same source data, determine the necessary structure in the data structure based on the frequency of the data structure in the same source data, and then determine the structural integrity signal of the target analysis data based on the necessary structure. The risk identification module is used to acquire target analysis data corresponding to structurally complete signals. Based on the data structure of the target analysis data, the data information corresponding to each data structure is marked as a structural data segment. Then, based on the same source data, the structural data segments are divided into common data and new data. Public data and new data are analyzed separately to determine the risk coefficients of public data and new data. Finally, based on the risk coefficients of public data and new data, the comprehensive risk value of the target analysis data is determined. Based on the comprehensive risk value, storage signals and security anomaly signals are then determined. The information storage module is used to store the target analysis data corresponding to the stored signal; The anomaly display module is used to display the target analysis data corresponding to the safety anomaly signal on the display terminal, and at the same time generate audio and visual reminder information.

2. The data storage and information security processing system based on cloud computing according to claim 1, characterized in that, Methods for determining common source data include: The target analysis data is converted into text using a multimodal Transformer model. The text data is then split into words, and the resulting content words are marked as keywords. These keywords are then used as search terms, and a large-scale data search is performed in the storage database to obtain the search data. The retrieved data is sequentially compared with the target analysis data to calculate the similarity value. The information similarity value is then compared with the information similarity threshold A1. If the information similarity value is greater than or equal to the information similarity threshold A1, the corresponding retrieved data is marked as source data. Conversely, if the information similarity value is less than the information similarity threshold A1, the corresponding retrieved data is marked as irrelevant information.

3. The data storage and information security processing system based on cloud computing according to claim 1, characterized in that, Methods for determining structurally intact signals include: Acquire all data from the same source, analyze the data scenario according to the target data, and identify the data structure in each piece of data from the same source. Here, the data scenario refers to the specific business background of data generation, flow, and application, and the data structure is the organization form of information in the data scenario. All data structures are sequentially labeled as target structures. Source data is obtained, and the existence of target structures in source data is identified. If target structures exist in source data, a presence label is set for this source data. All source data is traversed, and the number of source data containing the presence label is counted and marked as the structure occurrence frequency TC. The structure occurrence frequency FC of the target structure is obtained by using the formula TC÷ZH=FC, where ZH represents the total number of single source data. The frequency of occurrence of the structure FC is compared with the frequency threshold X1. If FC < X1, the target structure is marked as an accidental structure; otherwise, if FC ≥ X1, the target structure is marked as a necessary structure. Extract all necessary structures from the data structure, then obtain the target analysis data, identify the data structure in the target analysis data. If all necessary structures exist in the data structure of the target analysis data, a complete structure signal is generated. Conversely, if one or more necessary structures do not appear in the data structure of the target analysis data, a missing structure signal is generated.

4. The data storage and information security processing system based on cloud computing according to claim 3, characterized in that, When a structural missing signal is generated, the corresponding target analysis data is transmitted to the anomaly display module. The anomaly display module displays the target analysis data corresponding to the structural anomaly signal on the display terminal and generates an audio-visual reminder message. The corresponding data management personnel then re-confirm the safety and structural integrity of the target analysis data.

5. The data storage and information security processing system based on cloud computing according to claim 1, characterized in that, The risk identification module includes a data processing unit, a common analysis unit, a difference analysis unit, and a comprehensive processing unit; The data processing unit is used to acquire target analysis data corresponding to the structurally complete signal. Based on the data structure of the target analysis data, the data information corresponding to each data structure is marked as a structural data segment. Then, based on the same source data, the structural data segments are divided into common data and new data. The public analysis unit is used to analyze public data based on the frequency of public data occurrence in the same source data and the number of data anomalies in the data source location of the public data, and to determine the risk coefficient of the public data. The difference analysis unit is used to identify the data source location of the new data, extract historical output data from this data source location, analyze the historical output data, and determine the risk coefficient of the new data. The integrated processing unit is used to determine the comprehensive risk value of the target analysis data based on the risk coefficients of public data and new data, and then determine the storage signal and security anomaly signal based on the comprehensive risk value.

6. The data storage and information security processing system based on cloud computing according to claim 5, characterized in that, Methods for identifying public data and new types of data include: Obtain target analysis data and source data, divide the target analysis data according to data structure, and mark the data information corresponding to each data structure as a structure data segment; The structured data segment is searched in the same source information. If the structured data segment exists in the same source information, it is marked as public data. Otherwise, if the structured data segment does not exist in the same source information, it is marked as new data.

7. A cloud computing-based data storage information security processing system according to claim 5, characterized in that, Methods for determining the risk coefficient of novel data include: Select any common data b, identify the number of times common data b appears in the same source data, and mark it as the number of segments. Divide the number of segments of common data b by the total number of the same source data, and mark the result as the number of segments frequency FDb. Then identify the data source location of public data b, and set a time interval according to the periodic time, with the current time as the time node. The current time is the end time of the time interval, and the periodic time is the length of this time interval. Within a time interval, count the number of data anomalies in the data source locations of public data b, and divide this number of data anomalies by the number of data produced in this data source location to obtain the location anomaly frequency FWb. Using calculation formula Obtain the anomaly coefficient Yb of the public data b, where, This represents the error coefficient indicating the location of the data source. All public data are processed according to the above method to obtain the error coefficient of each public data. In the target analysis data, the error coefficients of all public data are obtained, and the mean of the error coefficients is calculated. The result of the mean calculation is marked as the risk coefficient Gd of the public data.

8. A cloud computing-based data storage information security processing system according to claim 5, characterized in that, Methods for determining the risk coefficient of novel data include: Based on the data source location of the new data, historical output data from this data source location is extracted. Then, the new data and historical output data are processed for similarity to obtain data similarity value. Historical output data with data similarity value greater than or equal to the data similarity threshold X2 are selected and marked as similar control data. Obtain all similar control data and perform mean processing on the similar control data. Mark the obtained mean result as the control characteristic value. Then, subtract the control characteristic value from the new data to obtain the offset value. Take the absolute value of the offset value and mark the absolute value as the offset representative value. The offset representative value of all new data is calculated according to the above method. Then, based on all the new data in the target analysis data, the offset representative value of the new data is averaged and the resulting value is marked as the risk coefficient Gx of the new data. When the data source of the new data has not generated any data before, the risk coefficient of the new data is directly set to m, that is, Gx=m.

9. A cloud computing-based data storage information security processing system according to claim 5, characterized in that, Methods for determining storage signals and security anomaly signals include: Obtain the risk coefficient Gd of the common data and the risk coefficient Gx of the novel data in the target analysis data, and then use the calculation formula. The comprehensive risk value Zf of the target analysis data is obtained, where, and These are the proportionality coefficients, and ; The comprehensive risk value Zf of the target analysis data is compared with the risk threshold K1. If Zf ≤ K1, a storage signal is generated and the target analysis data corresponding to the storage signal is transmitted to the information storage unit. Conversely, if Zf > K1, a safety anomaly signal is generated and transmitted to the anomaly display module.

10. A cloud computing-based data storage information security processing method, wherein the method is applied to a cloud computing-based data storage information security processing system as described in any one of claims 1-9, characterized in that, The method specifically includes the following steps: Step 1: Read the data to be stored and mark it as target analysis data. Convert the target analysis data into text form and set search terms. Based on the search terms, perform big data retrieval and processing in the storage database to identify data from the same source. Step 2: Based on the data scenario of the target analysis data, set the data structure, obtain the same source data, determine the necessary structure in the data structure based on the frequency of the data structure in the same source data, and then determine the structural integrity signal of the target analysis data based on the necessary structure. Step 3: Obtain the target analysis data corresponding to the structurally complete signal. Based on the data structure of the target analysis data, mark the data information corresponding to each data structure as a structural data segment. Then, based on the same source data, divide the structural data segments into common data and new data. Step 4: Based on the frequency of public data occurrences in the same source data and the number of data anomalies in the data source location of the public data, analyze the public data to determine the risk coefficient of the public data. Then, identify the data source location of the new data, extract the historical output data from this data source location, analyze the historical output data to determine the risk coefficient of the new data, and finally, based on the risk coefficients of the public data and the new data, determine the comprehensive risk value of the target analysis data. Based on the comprehensive risk value, determine the storage signal and security anomaly signal. Step 5: When a stored signal is detected, the target analysis data corresponding to the stored signal is stored in the data storage. When a security anomaly signal is detected, the corresponding target analysis data is displayed on the display terminal, and an audio-visual reminder message is generated at the same time.

Images