A cloud backup method and system for communication software

Abstract

The present invention relates to the field of data management technology, specifically a cloud-based backup method and system for communication software, comprising the following steps: collecting differentiated device access logs based on the cloud backup node to which the user is connected, extracting node jump sequences to identify repeated combinations, screening consistent combinations based on path sequence comparison data labels and assigning them to scheduling channels, extracting the revision time field of the communication record, arranging the field structure in accordance with the access sequence, and constructing a multi-level index node to generate a backup index tree configuration result. In the present invention, by collecting device access logs and extracting node jump sequences, repeated path behavior can be identified, effectively improving the accuracy of data organization and achieving session data time synchronization. A structured index path set can be arranged in accordance with the field access sequence, and multi-level index nodes can be constructed according to the path position to ensure the orderliness of data writing and retrieval efficiency, avoid data redundancy and backup conflicts, and enhance the reliability and query efficiency of data recovery.

Description

Technical Field

[0001] The present invention relates to the field of data management technology, and in particular to a cloud backup method and system for communication software. Background Art

[0002] The field of data management technology involves core issues such as the collection, storage, organization, maintenance, backup, recovery and access control of various types of data. It covers the entire data life cycle management technology, supports structured and unstructured data processing, access logic control, data consistency assurance, disaster recovery mechanism construction, data synchronization and version management, cloud storage scheduling and other aspects, and supports the security, integrity and availability of data in different systems or platforms.

[0003] Among them, the cloud backup method of traditional communication software refers to a solution for remote storage of user communication data to ensure data security. The technical issue it targets is how to effectively synchronize local message records, pictures, voice, video and other user data to a remote server for periodic incremental backup in a mobile communication software environment. After extracting data files by user-authorized access to the device's local storage directory, they are transmitted to the cloud server via the HTTPS protocol. At the same time, incremental identification is performed based on the timestamp comparison of database records and updated data is retrieved, and classified storage and index management are carried out with the help of object storage services.

[0004] The existing technology extracts data files by accessing the local storage directory and uses timestamp comparison for incremental identification, relying on a single time field to determine data updates. In scenarios where multiple devices are used in parallel, it is easy for the incremental identification to be inaccurate due to inconsistent time fields, resulting in data omissions or duplicate backups. In addition, the traditional method fails to effectively associate the device behavior path with the data item characteristics, resulting in loose organization of backup data and low retrieval efficiency. It is easy to cause data confusion or slow recovery in long-term backup and recovery, reducing the overall availability and reliability of the data backup system. Summary of the Invention

[0005] The purpose of the present invention is to solve the shortcomings of the prior art and to propose a cloud backup method and system for communication software.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: a cloud backup method for communication software, comprising the following steps:

[0007] S1: Based on the cloud backup nodes connected by users in the communication software, collect the access log data of the nodes on different devices, extract the node jump sequence in the continuous access behavior by device number, identify repeated node combinations, and obtain the device behavior path mapping table;

[0008] S2: Calling the device path in the device behavior path mapping table, detecting the client tags bound to the associated data items in order of the paths, comparing the tags of the data items with the nodes in the paths, selecting node data combinations with consistent behavior types and path order, and outputting a cloud backup scheduling node chain group;

[0009] S3: Calling the communication records of each channel in the cloud backup scheduling node chain group, extracting the device number, session identifier, and message sending time, categorizing the devices and locating the start node time field of the session, performing time comparison and uniformly revising the time field of the device records to obtain a unified timetable for the session data;

[0010] S4: Call the field information of each data in the unified time table of the session data, extract the sending time, message length, file format and session number fields, combine the access order of the corresponding fields, arrange the field structure according to the access order, and form a field sequence index path set.

[0011] As a further solution of the present invention, the device behavior path mapping table includes a path number, a node jump mode, and a device behavior label; the cloud backup scheduling node chain group includes a channel identifier, a node binding mapping, and a path sequence number; the session data unified time table includes a message time mapping table, a device unified timestamp, and a synchronization status identifier; and the field sequence index path set includes a field sequence list, a field position index, and an access structure label.

[0012] As a further solution of the present invention, the steps of obtaining the device behavior path mapping table are specifically as follows:

[0013] S111: Based on the cloud backup node connected by the user in the communication software, access log data on different devices is collected, the complete sequence of device numbers and corresponding access behaviors is extracted, and the log records are arranged in order of access to obtain a device access behavior sequence group;

[0014] S112: Calling the device access behavior sequence group, extracting the node number field in the access record for the continuous access behavior sequence under the same device number, comparing the node jump segments, identifying jump segments with the same node combination, and performing number mapping on the repeated jump segments to obtain a node jump combination set;

[0015] S113: Based on the node jump combination set, the frequency of occurrence of repeated combinations in the node jump path of each device is accumulated, and the mapping strength value of the device in the node combination path is calculated by combining the node jump path length and the position of the jump combination in the path. The mapping strength values ​​are summarized according to the device number, and the node behavior relationship of the device is analyzed to obtain the device behavior path mapping table.

[0016] As a further solution of the present invention, the steps for obtaining the cloud backup scheduling node chain group are specifically as follows:

[0017] S211: Calling the device path in the device behavior path mapping table, detecting the client tags bound to the associated data items in order of the paths, comparing the tags of the data items with the nodes in the paths, filtering the node data combinations based on the consistency between the data item tags and the node tags, and obtaining the node data combinations with consistent tags;

[0018] S212: Based on the node data combinations with consistent labels, screening node data combinations with consistent behavior types and path sequences, calculating behavior sequence matching values, screening node data combinations that meet the requirements according to the behavior sequence matching values, and generating sequence matching node data combinations;

[0019] S213: Based on the sequential matching node data combination and the continuous mapping relationship between nodes and data items, the node chain group that meets the mapping relationship is classified into the scheduling channel, and the cloud backup scheduling node chain group is output.

[0020] As a further solution of the present invention, the steps of obtaining the unified timetable of session data are specifically as follows:

[0021] S311: Based on the communication records of each channel in the cloud backup scheduling node chain group, extract the device number, session identifier and message sending time, classify the differentiated device numbers, and aggregate the communication records of the same device number to obtain a device communication record set;

[0022] S312: Call the device communication record set, locate the starting node time field of the session for each session identifier in the set, extract the starting node time field data, compare the starting node time under the differentiated device number, calculate the unified time difference measurement value between devices, uniformly revise the device record time field, combine the revised time field set, integrate the time record of each session identifier, and generate a unified time table for session data.

[0023] As a further solution of the present invention, the steps of obtaining the field sequence index path set are specifically as follows:

[0024] S411: Retrieving field information of multiple data items in the unified session data timetable, extracting the sending time, message length, file format, and session number fields recorded in each data item, and sequentially extracting corresponding fields based on the field identifier position for each data item to obtain a field combination set;

[0025] S412: Using the sending time, message length, file format and session number field contents in the field combination set, combined with the field access sequence information recorded in the client access log data, and comparing them with the field access sequence number under the same number in the access log, combined with the real-time arrangement sequence of each group of data fields, a field sequence index path set is obtained.

[0026] As a further embodiment of the present invention, the method further comprises the steps of:

[0027] S5: Call the path position corresponding to the field in the field sequence index path set, build multi-level index nodes in sequence, complete the upper, middle and bottom node configurations in the path sequence, bind each data write request to the corresponding structure path, and generate a backup index tree configuration result;

[0028] The backup index tree configuration result includes index node structure, data binding relationship, and structure level identification.

[0029] As a further solution of the present invention, the steps of obtaining the backup index tree configuration result are specifically as follows:

[0030] S511: Calling the path position corresponding to the field in the field sequence index path set, locating the position information of the field in the path set in sequence according to the field arrangement order, and combining the position information with the field name to generate a field path index sequence;

[0031] S512: Based on the field path index order, construct multi-level index nodes in sequence, aggregate fields with the same path position in the same level into the same node, set the structural relationship between the upper layer, middle layer and bottom layer respectively, and generate an index hierarchical node connection structure;

[0032] S513: Call the index level node connection structure, perform path identification processing on each data write request, bind the data content to the corresponding node position according to the field mapping relationship, construct a tree structure in sequence according to the structural hierarchy, and generate a backup index tree configuration result.

[0033] The cloud backup system for communication software is used to execute the cloud backup method for communication software, and the system includes:

[0034] The node access collection module extracts the device number, node address, and access time fields based on the cloud backup node connected by the user in the communication software, groups them by device number, arranges the node order by access time, determines the consecutive access node jumps, and generates a device behavior path mapping table;

[0035] The behavior path mapping module uses the device behavior path mapping table to identify the node combination in the jump sequence, counts the frequency of occurrence of the node combination, and combines the device number to obtain the cloud backup scheduling node chain group;

[0036] The node label screening module calls the cloud backup scheduling node chain group, extracts the client labels bound to the associated data items based on the path order, and generates a unified timetable for session data by comparing the node labels with the data item labels one by one;

[0037] The session time unification module calls the communication records of the channels in the session data unified time table, extracts the device number, session identifier, and message sending time, classifies the data according to the device number, locates the sending time of each session, and generates a field sequence index path set;

[0038] The backup index module calls the field sequence index path set, extracts the sending time, message length, file format, and session number, sorts the fields based on the sending time, extracts the field path position, builds index nodes in path order, binds data records, and obtains the backup index tree configuration result.

[0039] Compared with the prior art, the advantages and positive effects of the present invention are:

[0040] In the present invention, repeated path behaviors are identified by collecting device access logs and extracting node jump sequences. Node data combinations are screened based on the sequential consistency of device paths and data labels to form a continuously mapped backup scheduling chain, effectively improving the accuracy of data organization. The device record time field is uniformly revised with the help of time comparison of communication records to achieve session data time synchronization. The structured index path set is arranged in combination with the field access order, and multi-level index nodes are constructed according to the path position to ensure the orderliness of data writing and retrieval efficiency. The overall process improves the structural integrity of data backup and the consistency of access timing, avoids data redundancy and backup conflicts, and enhances the reliability of data recovery and query efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 It is a schematic diagram of the workflow of the present invention;

[0042] Figure 2 This is a flowchart for obtaining the device behavior path mapping table in the present invention;

[0043] Figure 3 This is a flowchart for obtaining a cloud backup scheduling node chain group in the present invention;

[0044] Figure 4 This is a flow chart for obtaining a unified timetable for session data in the present invention;

[0045] Figure 5This is a flowchart for obtaining a field sequence index path set in the present invention;

[0046] Figure 6 This is a flowchart for obtaining the backup index tree configuration results in the present invention. DETAILED DESCRIPTION

[0047] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0048] In the description of the present invention, it should be understood that the terms "length," "width," "up," "down," "front," "back," "left," "right," "vertical," "horizontal," "top," "bottom," "inside," "outside," and the like, indicating positions or relationships, are based on the positions or relationships shown in the accompanying drawings and are intended only to facilitate the description of the present invention and simplify the description. They do not indicate or imply that the devices or elements referred to must have a specific orientation, be constructed, or operate in a specific orientation. Therefore, they should not be construed as limiting the present invention. Furthermore, in the description of the present invention, "plurality" means two or more, unless otherwise expressly and specifically defined.

[0049] See also Figure 1 The present invention provides a technical solution, a cloud backup method for communication software, comprising the following steps:

[0050] S1: Based on the cloud backup nodes connected by users in the communication software, collect the access log data of the nodes on different devices, extract the node jump sequence in the continuous access behavior by device number, identify repeated node combinations, and obtain the device behavior path mapping table;

[0051] S2: Call the device path in the device behavior path mapping table, detect the client tags bound to the associated data items in order of the paths, compare the tags of the data items with the nodes in the paths, select the node data combinations with the consistent behavior type and path order, classify the combinations into the scheduling channel, and output the cloud backup scheduling node chain group based on the continuous mapping relationship between the nodes and data items in the scheduling channel;

[0052] S3: Call the communication records of each channel in the cloud backup scheduling node chain group, extract the device number, session identifier, and message sending time, classify them by device, and locate the time field of the starting node of the session. After performing time comparison, uniformly revise the time field of the device records to obtain a unified timetable of session data;

[0053] S4: Call the field information of each data in the unified time table of session data, extract the sending time, message length, file format and session number fields, combine the access order of the corresponding fields in the client access log data, arrange the field structure according to the access order, and form a field sequence index path set;

[0054] S5: Call the path position corresponding to the field in the field sequence index path set, build multi-level index nodes in sequence, complete the upper, middle and bottom node configurations in the path sequence, bind each data write request to the corresponding structure path, and generate the backup index tree configuration result;

[0055] The device behavior path mapping table includes the path number, node jump mode, and device behavior label. The cloud backup scheduling node chain group includes the channel identifier, node binding mapping, and path sequence number. The session data unified time table includes the message time mapping table, device unified timestamp, and synchronization status identifier. The field sequence index path set includes the field sequence list, field position index, and access structure label. The backup index tree configuration result includes the index node structure, data binding relationship, and structure hierarchy identifier.

[0056] See also Figure 2 , the specific steps for obtaining the device behavior path mapping table are:

[0057] S111: Based on the cloud backup node connected by the user in the communication software, access log data on different devices is collected, the complete sequence of device numbers and corresponding access behaviors is extracted, and the log records are arranged in order of access to obtain a device access behavior sequence group;

[0058] The access log data on differentiated devices is collected, and all bound terminal devices are located through user identification. A unique number index is established for each device in turn, and devices A, B, and C are marked as ID001, ID002, and ID003 respectively. The log entries corresponding to each number are then extracted one by one from the cloud node database. The extracted fields include: access timestamp, target node number, access type, and authentication status. The timestamp is accurate to milliseconds, the access type is GET / POST, and the authentication status is recorded using a Boolean variable. The log entries are classified by device number and sorted in ascending order according to the timestamp field to obtain the continuous access records of each device in the order of actual operation. Regarding the behavior sequence, the access records generated by device ID001 within a certain period of time are as follows: node N001 accessed at 08:12:23.345, node N002 accessed at 08:12:24.182, and node N003 accessed at 08:12:27.981. The three access behaviors constitute the access behavior sequence {N001→N002→N003} in chronological order. To prevent data anomalies from interfering with the behavior path, entries with an authentication status of False are removed from the access behavior, and only valid access behaviors after successful user authentication are retained. After removing invalid records, the node integrity of each access sequence is reconfirmed, and the device number is used as the index item to obtain the device access behavior sequence group.

[0059] S112: Calling the device access behavior sequence group, extracting the node number field in the access record for the continuous access behavior sequence under the same device number, comparing the node jump segments, identifying jump segments with the same node combination, and performing number mapping on the repeated jump segments to obtain a node jump combination set;

[0060] For each device access sequence in the set, the node number field in the access behavior record is extracted, and a jump path list is formed in the access order. The node combination of each jump path is compared in different devices. Set device ID001 to have a path sequence {N001→N002→N003→N004}, and ID002 to {N001→N002→N003→N005}. There is a repeated node combination {N001→N002→N003} between the two. Then this combination can be marked as a repeated path segment. Determine whether the repeated combination has a continuous node relationship with a path segment length greater than or equal to 3. If the condition is met, extract the segment combination to establish a standard path index table, map the device path, and mark the path that meets this feature. The repeated segments of the feature are compared and numbered, such as {N001→N002→N003} is assigned the path segment label R01, and the starting position and ending position of the path segment in the original access sequence are recorded. The access path can be converted into a logical jump path sequence composed of a mixture of ordinary nodes and combination labels. The repeated combination validity threshold is set to an occurrence frequency ≥ 2 in the path segment judgment. If the number of occurrences of a certain node combination in different device paths is lower than this threshold, it is not considered a repeated path segment. If the combination {N002→N004} only appears once in ID001, it is excluded from the valid node combination. The combination numbers and structures that meet the path segment length and repetition frequency requirements are constructed in a list form to obtain a node jump combination set.

[0061] S113: Based on the node jump combination set, the frequency of repeated combinations in the node jump path of each device is accumulated, and the formula is used to combine the length of the node jump path and the position of the jump combination in the path:

[0062] ;

[0063] Calculate the mapping strength value of the device in the node combination path, summarize the mapping strength values ​​by device number, analyze the node behavior relationship of the device, and obtain the device behavior path mapping table;

[0064] in, Representative equipment The node behavior mapping strength value of Representative equipment Middle The frequency of occurrence of node combinations, Representative equipment Middle The path length of the node combination, Representative equipment Middle The jump position number in the path of the node combination, Representative equipment Middle The jump density value of the node combination, is the number of node combinations in the device;

[0065] Formula calculation logic: formula is used to calculate the device Node behavior mapping strength value , its core logic consists of two parts. The first part is the product of the frequency term and the structure term, that is, , which reflects the product of the number of times a node combination appears in the path and the length of the structure, divided by the relative position of the combination in the path The square root of is used to weaken the influence of subsequent combinations in the path and form an attenuation mechanism; the second part is the jump density value of the node combination in the device The average value of the two calculation results represents the density of jumps in the path. The difference between the two calculation results is taken as the absolute value. This ensures that the degree of mapping anomaly can be measured regardless of whether the frequency structure is high or the jump density is low. Overall, the formula forms a composite scoring mechanism by integrating four parameters: structure frequency, length, location, and jump density. It quantitatively evaluates the structural stability and access consistency of the path combination. It is suitable for comparing and classifying the path structures of multiple devices under the same logical service.

[0066] The node behavior mapping strength value is a composite indicator that measures the structural characteristics of repeated node combinations and jump distribution in a device's access path. By calculating the frequency of occurrence, path length, path location, and jump density of node combinations, the structural characteristics of their behavior patterns are extracted. A higher value indicates a denser and more clearly structured repeated jump combination pattern in the device's access path, facilitating path normalization mapping and identifying abnormal behavior.

[0067] Calculate the reciprocal mean of the jump times between adjacent combination segments in the same device and set it as the jump density value , if the total number of jumps in the path is 12, and the spacing between repeated combination segments is 2, 2, 4, and 4 respectively, then the jump density is , substitute the above four parameters into the formula:

[0068] Take device ID001 as an example. It has three repeating path combinations: R01, R03, and R05. The parameter settings are as follows:

[0069] Table 1 Equipment parameters

[0070]

[0071] As shown in Table 1, the calculation process is as follows:

[0072] Calculate the product of each term and divide by the square root: R01: ;R03:

[0073] R05: ; Calculate the sum and subtract the mean jump density:

[0074] Total transfer items:

[0075] Average density term:

[0076] Jump Strength Value:

[0077] ;

[0078] The proceeds Record the node behavior mapping value of ID001, build the node behavior map of each device in sequence according to the device number, and generate the device behavior path mapping table;

[0079] Table 2 shows the example parameters used in the above calculation process:

[0080] Table 2 Device node combination statistical parameters

[0081]

[0082] As shown in Table 2, different node combination items have clear structured records in the device path, which facilitates the subsequent unified structure scoring and difference quantification;

[0083] The results show that by calculating the node behavior mapping strength value, a unique identification and structural characterization parameter can be provided for the device behavior path mapping table, forming a device path structure description result that is comparable and decidable.

[0084] See also Figure 3 ,The specific steps for obtaining the cloud backup scheduling node chain group are:

[0085] S211: Calling the device path in the device behavior path mapping table, detecting the client tags bound to the associated data items in order of the paths, comparing the tags of the data items with the nodes in the paths, filtering the node data combinations based on the consistency between the data item tags and the node tags, and obtaining the node data combinations with consistent tags;

[0086] Based on the path information extracted from the device behavior path mapping table, the device identification information of the nodes in the path is obtained. The device identification information includes the device ID, the node sequence index and the bound client tag. The associated data item is obtained. The bound client tag and the data item code are recorded in the data item. The client tag bound to each data item is detected in sequence by traversal. For each node, its device identification information is called and compared one by one with the client tag in the data item. Matching judgment is adopted. If the client tag ID of the node is consistent with the client tag ID of the data item, the corresponding relationship between the node and the data item is recorded, and the nodes and data items that meet the conditions are combined into a preliminary node data combination. In the example , set the client tag IDs bound to nodes A, B, and C in the path to T1, T2, and T3 respectively, and the client tag IDs bound to data items D1, D2, and D3 to T1, T3, and T2 respectively. Then, through comparison, node A corresponds to data item D1, node B corresponds to data item D3, and node C corresponds to data item D2, forming three groups of node data combinations respectively. During the detection process, the nodes are processed in ascending order of the sequence index number to avoid sequence confusion. In the comparison process, strict equivalence judgment is adopted. The client tag IDs must be completely consistent and fuzzy matching is not supported. If the client tag ID adopts the 16-bit UUID encoding method, it should be compared bit by bit. The preliminary node data combination can be listed in the following table:

[0087] Table 3 Comparison of node and data item labels

[0088]

[0089] As shown in Table 3, by comparing the client tag IDs of the nodes and data items, the node data combination with consistent tags is determined, and the node data combination with consistent tags is obtained.

[0090] S212: Based on the node data combinations with consistent labels, filter the node data combinations with consistent behavior types and path sequences using the formula:

[0091] ;

[0092] Calculate the behavior sequence matching value, filter the node data combination that meets the requirements according to the behavior sequence matching value, and generate the sequence matching node data combination;

[0093] in, is the behavior sequence matching value, Represents the behavior type code of the node binding, Represents the sequential index value of the node in the path, Represents the timestamp of the data item, Represents the client tag ID, Represents the node label ID, Represents the number of nodes in the path, Represents the number of node combinations with consistent labels;

[0094] The calculation logic of the formula is as follows: Encode the behavior type of the node binding , sequential index value , data item timestamp Sum them up separately to form a unified sum of behavioral time features, reflecting the overall distribution characteristics of the nodes in the path in terms of behavior and time; respectively, the data item client label set and node client label collection Perform difference calculation, the difference reflects the consistency between the path node and the data item in the label dimension. The smaller the difference, the higher the label matching degree. Multiply the sum of the behavior time features with the label difference result, and combine it with the number of path nodes. Number of node combinations consistent with the label The sum of is normalized, and after normalization, the absolute value is taken and square rooted to output the behavior sequence matching value. , eliminate the influence of a single data dimension on the overall matching degree, ensure the dimensionality of each participating quantity is unified, and the calculation results reflect the overall matching degree of the node path in terms of behavioral characteristics and data consistency;

[0095] The behavior sequence matching value is used to measure the consistency of the path sequence and time characteristics of the device node based on the consistency of the behavior type and data item label. This value integrates the node behavior code, sequence index, timestamp and label matching, and quantifies the overall matching relationship of the node combination through a unified normalized operation. The closer the behavior sequence matching value is to 0, the higher the consistency of the node data in behavior and sequence, and vice versa.

[0096] Extract the behavior type code and sequential index value bound to each node, extract the timestamp and client tag ID of the data item, and then construct the data set separately 、 、 、 、 、 、 ,in Represents a node behavior type encoding set, Represents a set of sequential indexes of path nodes, Represents a collection of data item timestamps, Represents a client tag ID set, Represents a set of node label IDs, Indicates the number of nodes in the path, Indicates the number of node combinations with consistent labels;

[0097] Set the behavior type codes to 101, 102, and 103, the order indexes to 1, 2, and 3, the timestamps to 1620000001, 1620000020, and 1620000030, the client label ID and the node label ID are equal, and the values ​​are 1001, 1002, and 1003, respectively. , the number of consistent label combinations , the specific calculation process is as follows:

[0098] ;

[0099] ;

[0100] ;

[0101] ;

[0102] In the formula, Represents the sum of the node behavior type code, sequence index and data item timestamp, Represents the sum of the differences between the client label ID set and the node label ID set. The sum of the total number of nodes and data items, calculated from the above, is the behavior order matching value. ,Since the value is zero, it does not meet the subsequent screening requirements, so it is necessary to introduce interference value adjustment. In actual application, there is an error adjustment item in the behavior coding, and the resampling code is 101, 103, and 105. After updating, recalculate:

[0103] ;

[0104] ;

[0105] If the interference item is still zero after adjustment, a benchmark value needs to be set for secondary screening. The benchmark value is set to 0.1, which is taken from the standard deviation of the mean of the device operation delay. Through the actual measurement of multiple groups of samples, the standard deviation of the device delay is about 0.08 to 0.12 seconds. Therefore, the benchmark value is set within the range and screening is performed according to the benchmark value. , then remove the combination and obtain the sequential matching node data combination through screening.

[0106] S213: Based on the sequential matching of node data combinations and the continuous mapping relationship between nodes and data items, the node chain groups that meet the mapping relationship are classified into the scheduling channel, and the cloud backup scheduling node chain group is output;

[0107] Call the mapping relationship between nodes and data items to extract continuous node chain groups. Continuous node chain groups refer to node IDs that are consecutive in sequential index, and the difference between the timestamps of the corresponding data items does not exceed the set time limit. The set time limit is 5 seconds, with 10 times the value of the device acquisition cycle of 0.5 seconds as the upper limit. Set the node order and timestamps as follows:

[0108] Table 4 Node chain group timestamp table

[0109]

[0110] As shown in Table 4, the time stamp difference between node B and node A is 4 seconds, which is less than 5 seconds. The time stamp difference between node C and node B is 5 seconds, which is equal to the time limit. Both meet the continuity requirements. Nodes A, B, and C are combined into a chain group in sequence and classified into the scheduling channel to generate a cloud backup scheduling node chain group.

[0111] See also Figure 4 ,The steps for obtaining the unified timetable of session data are as follows:

[0112] S311: Based on the communication records of each channel in the cloud backup scheduling node chain group, extract the device number, session identifier, and message sending time, classify the differentiated device numbers, and aggregate the communication records of the same device number to obtain a device communication record set;

[0113] Communication record data is collected, and the corresponding device number, session identifier, and message sending time fields are extracted one by one. Data classification is performed on the device number field in the communication record data, that is, communication records with the same device number are grouped according to the device number and classified into a set of device communication records. Taking device number A123 as an example, after classification, the corresponding set contains several communication records, and the recorded message sending times are 2025-05-01 12:00:00, 2025-05-0112:01:30, 2025-05-01 12:03:00, etc. By sequentially scanning the device number field in each communication record, and gathering records with the same device number into one set, in the case of large data volume, a doubly linked list can be used to store the device communication record set to improve the efficiency of subsequent traversal. For device number A124, the sending times in the set obtained after classification are 2025-05-01 12:00:10, 2025-05-01 12:01:40, 2025-05-01 12:03:20, etc. After the classification is completed, the device communication record sets are stored in independent groups according to the device numbers, which is convenient for subsequent positioning and processing of the session identifier to obtain the device communication record sets.

[0114] S312: Call the device communication record collection, locate the start node time field of each session identifier in the collection, extract the start node time field data, and compare the start node time under the differentiated device number using the formula:

[0115] ;

[0116] Calculate the unified time difference measurement value between devices, uniformly revise the device record time field, combine the revised time field set, integrate the time record of each session identifier, and generate a unified timetable for session data;

[0117] in, Representative equipment With equipment The unified time difference measure of Representative equipment The starting node time field value of Representative equipment The starting node time field value;

[0118] Formula calculation logic: based on the device With equipment The time field value of the starting node is calculated, and the time difference between the two is calculated by the absolute value operator. Ensure that the time difference is non-negative to avoid the time sequence affecting the stability of the results. Calculate the product of the two time fields and take the square root. It reflects the closeness of the magnitude of the time fields of the two devices. The nonlinear deviation introduced by the magnitude difference is eliminated by square root processing. Then the absolute value difference and the square root value are added together to make the time difference and the time product magnitude be measured on a unified scale. Using the denominator as the denominator, normalization is performed to scale the total amount of quantified time differences, keeping the metric value within a relatively stable range. This process comprehensively considers the differences and overall magnitude of the time fields of the two devices, ensuring that the unified time difference metric is both sensitive and robust. This is suitable for scenarios where time records on different devices have slight offsets but consistent magnitudes, further improving the consistency and accuracy of time revisions.

[0119] The unified time difference metric is a standardized value used to measure the relationship between the start node time difference and the overall magnitude of two devices in the same session. This metric combines the absolute value of the time difference with the square root of the time product and normalizes it to reflect the offset and consistency level of the time fields of the two devices. The closer the metric value is to 0, the closer the start times of the two devices are. The closer it is to 1, the larger the time deviation is, making it suitable for use as a basis for unified revision processing.

[0120] Traverse the records in the collection and group them by session ID. In each group, select the record with the earliest message sending time field as the start node time field of the session. For example, the start time field of session C1 with device number A123 is 2025-05-01 12:00:00, and the start time field of session C1 with device number A124 is 2025-05-01 12:00:10. Extract the start node time field data of each device and perform a pairwise comparison operation on the time fields.

[0121] Set the starting node time of device A123 43200 seconds, the starting node time of device A124 =43210 seconds, which can be calculated by substituting into the formula as follows:

[0122] ;

[0123] ;

[0124] ;

[0125] ;

[0126] ;

[0127] The unified time difference measurement value between devices is 0.4999. Furthermore, the unified revision threshold of the time difference is set to 0.5. If the value is lower than the threshold, a unified revision is performed. The revision method is to take the average value of the start node time of each device, that is: ; Revise the start time of each device to 43205 seconds, and the revised time is 12:00:05. Based on the revised time field set, integrate the time records of each session identifier, generate a time record table in a unified format, and form a unified time table for session data.

[0128] See also Figure 5 , the steps for obtaining the field sequence index path set are as follows:

[0129] S411: Retrieving the field information of multiple data items in the unified session data timetable, extracting the sending time, message length, file format, and session number fields recorded in each data item, and extracting the corresponding fields in sequence according to the field identifier position for each data item to obtain a field combination set;

[0130] Each data in the timetable is extracted in segments. The calling operation should be completed based on the field identification structure, which contains the field name, field type and offset position value of the field. By reading the structure definition, the actual starting offset position and length value of the four fields such as sending time, message length, file format and session number in each data can be located, and the content of each field is parsed and extracted in turn. The sending time can be recorded in Unix timestamp format, and the starting offset value of a certain data is set to 0 to record the 4-byte Unix timestamp 1617273600, which corresponds to 0:00 on April 1, 2021 Beijing time; the message length field follows immediately. If it is set to a 2-byte unsigned integer, it can be parsed as a message containing byte values ​​such as 128 bytes; the file format field can be represented by fixed characters, such as ".txt" or ".doc", in the original The record is stored in ASCII code, with a corresponding offset value of 6 and a length of 4 bytes. After reading, it is converted to character form. The session number field is set to a unique integer identifier, such as session number 10086, which is set as a 4-byte integer field starting at offset 10. After the above fields are extracted in sequence, the four fields are concatenated in the order of "sending time-message length-file format-session number" based on the data structure to form a field group. The extraction process is repeated for each record in the unified timetable. The following field combination group is formed for three consecutive data items: {1617273600, 128, ".txt", 10086}, {1617277200, 256, ".pdf", 10087}, {1617280800, 64, ".doc", 10088}, which are aggregated to form a field combination set.

[0131] S412: Using the sending time, message length, file format, and session number field contents in the field combination set, combined with the field access sequence information recorded in the client access log data, the data is compared with the field access sequence numbers under the same number in the access log, and combined with the real-time arrangement sequence of each group of data fields to obtain a field sequence index path set;

[0132] According to the structure of the field combination set, the values ​​of sending time, message length, file format and session number in each group are read one by one, and the session number is matched with the corresponding record in the client access log data. By parsing the field access sequence number field in the client access log, the order in which the fields corresponding to the session number are accessed on the client is identified. The field access sequence of the session number 10086 in the client access log is set to {3, 1, 4, 2}, which means that the third field (i.e., file format) in the original field order is accessed first, followed by the first field (i.e., sending time), the fourth field (session number) and the second field (message length). The field sequence in the field combination set needs to be reordered and arranged according to this order. The structure of the current field combination group, that is, the original field group {1617273600, 128, ".txt", 10086} is rearranged into {".txt", 1617273600, 10086, 128}; repeat the matching and rearrangement operations, perform log access order parsing and field rearrangement operations on each field group in the field combination set, and construct a new structure set consisting of the rearranged field sequence; the field sequence in each rearranged structure needs to be structurally encoded, using the first letter of the field type to express the field type, setting S to represent the sending time, L to represent the message length, F to represent the file format, and C to represent the session number. The aforementioned rearranged sequence is encoded as "FSCL" to construct a field sequence index path set.

[0133] See also Figure 6 The specific steps for obtaining the backup index tree configuration results are as follows:

[0134] S511: Call the path position corresponding to the field in the field sequence index path set, locate the position information of the field in the path set in sequence according to the field arrangement order, and combine the position information with the field name to generate the field path index sequence;

[0135] Extract each field in the data structure and its arrangement order information in the structure path set. When processing the data record of each customer in the enterprise customer information, the fields include customer number, contact information, order status, etc., which exist as nested objects in the original structure. Set the fields "customer.id", "customer.contact.phone", and "customer.orders.orderID". At this time, the field paths contained in the path set need to be extracted according to their hierarchy and order. Through the preset path extraction, read the position information of the fields in the structure in turn, clarify the nesting level of each field and the position number under the level, and compare the obtained path position with the field name. Perform binding combinations to form a field path sequence with position information. Set the field customer.contact.phone to path level 2 and sequence 3, which is represented as (2, 3, customer.contact.phone). Rearrange the above structural sequence according to the initial definition order of the fields, and output a set of field paths arranged in sequence. During the processing process, if a field is found to have multiple levels or repeated path names in the path, its path needs to be extended and set by appending a level prefix to indicate the unique position to avoid field index conflicts. The execution process outputs a set of structured field path index sequence data, which serves as the basis for subsequent node construction and generates a field path index sequence.

[0136] S512: Based on the field path index order, construct multi-level index nodes in sequence, aggregate fields with the same path position in the same level into the same node, set the structural relationship between the upper layer, middle layer and bottom layer respectively, and generate the index hierarchical node connection structure;

[0137] It is necessary to first classify the levels of the fields in the path, and group the fields with the same path level together. The field path sequence set in financial transactions includes fields such as account.id, account.details.balance, account.details.limit, transaction.id, and transaction.amount. "account.id" is a first-level field, "account.details.balance" and "account.details.limit" are second-level fields, and "transaction" related fields are another level field. In the index structure construction, the fields at the same level need to be aggregated to form nodes, and the relationship between the upper and lower nodes is established according to the field path index order. The path prefix is ​​marked as a connection clue to form the upper layer as " The nodes are bound by the path prefix and index order. For example, if there are two fields, "account.details.limit" and "account.details.threshold", both of which belong to the middle-level node "account.details", they should be set as its subordinate bottom-level nodes. During the process, each node should be marked with the number of fields it contains, the path name and its index position in the data to form a clearly structured node chain structure. After the node is set, each node is given a path identification identifier and its subordinate relationship chain is recorded for path positioning and structure binding when subsequent data is nested and written. After the execution process is completed, a clear and callable node structure set is output to generate an index-level node connection structure.

[0138] S513: Call the index level node connection structure, perform path identification processing on each data write request, bind the data content to the corresponding node position according to the field mapping relationship, and construct a tree structure in sequence according to the structural hierarchy to generate a backup index tree configuration result;

[0139] First, extract the field value of each record from the actual business data. Taking medical imaging as an example, the fields include patient.id, exam.date, exam.findings.description, etc. After receiving a data write request, the field path is identified field by field according to the index node structure built previously. According to the field path mapping relationship, the field value is extracted from the record. For example, if the field exam.findings.description is identified, the corresponding structure path is "description" under "findings" under the node "exam". According to the structure path, the field value is read from the data, such as "The nodule image is clear and can be seen". "See", and bind it to the structure path node. After the field completes the path binding, according to the constructed multi-level structural relationship, the structure corresponding to the field is nested layer by layer to form a complete tree structure. During the nesting process, if the field value is not obtained under a path node, the empty value node structure is retained to maintain structural consistency. After the entire tree structure is constructed, it is stored in the form of a configuration file or a structured document, and index registration is performed for data retrieval. The fields written into the structure all have a path label and content binding relationship, which facilitates subsequent operations to quickly locate and read. After the operation process is completed, the written fields are integrated into the complete structure by node, and the tree structure configuration entity that can be called is output to generate a backup index tree configuration result.

[0140] The cloud backup system for communication software is used to execute the cloud backup method for communication software described above, and the system includes:

[0141] The node access collection module extracts the device number, node address, and access time fields based on the cloud backup node connected by the user in the communication software, groups them by device number, arranges the node order by access time, determines the consecutive access node jumps, and generates a device behavior path mapping table;

[0142] The behavior path mapping module uses the device behavior path mapping table to identify the node combination in the jump sequence, count the frequency of node combination occurrence, and combine it with the device number to obtain the cloud backup scheduling node chain group;

[0143] The node label filtering module calls the cloud backup scheduling node chain group, extracts the client labels bound to the associated data items based on the path order, and generates a unified timetable for session data by comparing the node labels with the data item labels one by one.

[0144] The session time unification module calls the communication records of the channels in the session data unified time table, extracts the device number, session identifier, and message sending time, categorizes the data according to the device number, locates the sending time of each session, and generates a field sequence index path set;

[0145] The backup index module calls the field sequence index path set, extracts the sending time, message length, file format, and session number, sorts the fields based on the sending time, extracts the field path position, builds index nodes in path order, binds data records, and obtains the backup index tree configuration result.

[0146] The above are merely preferred embodiments of the present invention and do not limit the present invention in any other form. Any technician familiar with the profession may use the technical content disclosed above to change or modify it into an equivalent embodiment with equivalent changes and apply it to other fields. However, any simple modification, equivalent change and modification made to the above embodiment based on the technical essence of the present invention without departing from the content of the technical solution of the present invention shall still fall within the scope of protection of the technical solution of the present invention.

Claims

1. A cloud backup method for communication software, characterized in that: The following steps are involved: S1: Based on the cloud backup nodes connected by users in the communication software, collect the access log data of the nodes on different devices, extract the node jump sequence in the continuous access behavior by device number, identify repeated node combinations, and obtain the device behavior path mapping table; S2: Calling the device path in the device behavior path mapping table, detecting the client tags bound to the associated data items in order of the paths, comparing the tags of the data items with the nodes in the paths, selecting node data combinations with consistent behavior types and path order, and outputting a cloud backup scheduling node chain group; S3: Calling the communication records of each channel in the cloud backup scheduling node chain group, extracting the device number, session identifier, and message sending time, categorizing the devices and locating the start node time field of the session, performing time comparison and uniformly revising the time field of the device records to obtain a unified timetable for the session data; S4: Calling the field information of each data in the unified timetable of the session data, extracting the sending time, message length, file format and session number fields, combining the access order of the corresponding fields, and arranging the field structure according to the access order to form a field sequence index path set; S5: Call the path position corresponding to the field in the field sequence index path set, build multi-level index nodes in sequence, complete the upper, middle and bottom node configurations in the path sequence, bind each data write request to the corresponding structure path, and generate a backup index tree configuration result; The backup index tree configuration result includes index node structure, data binding relationship, and structure level identification.

2. The cloud backup method for communication software according to claim 1, characterized in that: The device behavior path mapping table includes a path number, a node jump mode, and a device behavior tag; the cloud backup scheduling node chain group includes a channel identifier, a node binding mapping, and a path sequence number; the session data unified time table includes a message time mapping table, a device unified timestamp, and a synchronization status identifier; and the field sequence index path set includes a field sequence list, a field position index, and an access structure tag.

3. The cloud backup method for communication software according to claim 1, characterized in that: The steps for obtaining the device behavior path mapping table are specifically as follows: S111: Based on the cloud backup node connected by the user in the communication software, access log data on different devices is collected, the complete sequence of device numbers and corresponding access behaviors is extracted, and the log records are arranged in order of access to obtain a device access behavior sequence group; S112: Calling the device access behavior sequence group, extracting the node number field in the access record for the continuous access behavior sequence under the same device number, comparing the node jump segments, identifying jump segments with the same node combination, and performing number mapping on the repeated jump segments to obtain a node jump combination set; S113: Based on the node jump combination set, the frequency of occurrence of repeated combinations in the node jump path of each device is accumulated, and the mapping strength value of the device in the node combination path is calculated by combining the node jump path length and the position of the jump combination in the path. The mapping strength values ​​are summarized according to the device number, and the node behavior relationship of the device is analyzed to obtain the device behavior path mapping table.

4. The cloud backup method for communication software according to claim 3, characterized in that: The steps for obtaining the cloud backup scheduling node chain group are specifically as follows: S211: Calling the device path in the device behavior path mapping table, detecting the client tags bound to the associated data items in order of the paths, comparing the tags of the data items with the nodes in the paths, filtering the node data combinations based on the consistency between the data item tags and the node tags, and obtaining the node data combinations with consistent tags; S212: Based on the node data combinations with consistent labels, screening node data combinations with consistent behavior types and path sequences, calculating behavior sequence matching values, screening node data combinations that meet the requirements according to the behavior sequence matching values, and generating sequence matching node data combinations; S213: Based on the sequential matching node data combination and the continuous mapping relationship between nodes and data items, the node chain group that meets the mapping relationship is classified into the scheduling channel, and the cloud backup scheduling node chain group is output.

5. The cloud backup method for communication software according to claim 4, characterized in that: The steps for obtaining the unified timetable of the session data are specifically as follows: S311: Based on the communication records of each channel in the cloud backup scheduling node chain group, extract the device number, session identifier and message sending time, classify the differentiated device numbers, and aggregate the communication records of the same device number to obtain a device communication record set; S312: Call the device communication record set, locate the starting node time field of the session for each session identifier in the set, extract the starting node time field data, compare the starting node time under the differentiated device number, calculate the unified time difference measurement value between devices, uniformly revise the device record time field, combine the revised time field set, integrate the time record of each session identifier, and generate a unified time table for session data.

6. The cloud backup method for communication software according to claim 5, characterized in that: The specific steps for obtaining the field sequence index path set are: S411: Retrieving field information of multiple data items in the unified session data timetable, extracting the sending time, message length, file format, and session number fields recorded in each data item, and sequentially extracting corresponding fields based on the field identifier position for each data item to obtain a field combination set; S412: Using the sending time, message length, file format and session number field contents in the field combination set, combined with the field access sequence information recorded in the client access log data, and comparing them with the field access sequence number under the same number in the access log, combined with the real-time arrangement sequence of each group of data fields, a field sequence index path set is obtained.

7. The cloud backup method for communication software according to claim 6, characterized in that: The steps for obtaining the backup index tree configuration result are specifically as follows: S511: Calling the path position corresponding to the field in the field sequence index path set, locating the position information of the field in the path set in sequence according to the field arrangement order, and combining the position information with the field name to generate a field path index sequence; S512: Based on the field path index order, construct multi-level index nodes in sequence, aggregate fields with the same path position in the same level into the same node, set the structural relationship between the upper layer, middle layer and bottom layer respectively, and generate an index hierarchical node connection structure; S513: Call the index level node connection structure, perform path identification processing on each data write request, bind the data content to the corresponding node position according to the field mapping relationship, construct a tree structure in sequence according to the structural hierarchy, and generate a backup index tree configuration result.

8. A cloud backup system for communication software, characterized in that: The system is used to implement the cloud backup method of the communication software according to any one of claims 1 to 7, and the system includes: The node access collection module extracts the device number, node address, and access time fields based on the cloud backup node connected by the user in the communication software, groups them by device number, arranges the node order by access time, determines the consecutive access node jumps, and generates a device behavior path mapping table; The behavior path mapping module uses the device behavior path mapping table to identify the node combination in the jump sequence, counts the frequency of occurrence of the node combination, and combines the device number to obtain the cloud backup scheduling node chain group; The node label screening module calls the cloud backup scheduling node chain group, extracts the client labels bound to the associated data items based on the path order, and generates a unified timetable for session data by comparing the node labels with the data item labels one by one; The session time unification module calls the communication records of the channels in the session data unified time table, extracts the device number, session identifier, and message sending time, classifies the data according to the device number, locates the sending time of each session, and generates a field sequence index path set; The backup index module calls the field sequence index path set, extracts the sending time, message length, file format, and session number, sorts the fields based on the sending time, extracts the field path position, builds index nodes in path order, binds data records, and obtains the backup index tree configuration result.

Images