Enterprise cross-department process collaborative management system based on big data resources

Abstract

This invention discloses an enterprise cross-departmental process collaboration management system based on big data resources, comprising: an enterprise acquisition module: collecting raw process log data generated by multiple independent systems within the enterprise, performing unified semantic parsing preprocessing on the raw process log data, and generating cross-departmental process event streams; a data analysis module: performing cross-departmental temporal correlation pattern mining based on the cross-departmental process event streams, identifying effective cross-departmental event sequences, and constructing a cross-departmental collaboration knowledge graph based on process events in the cross-departmental event sequences as nodes and resource transfer relationships between events as edges; and a decision collaboration module: monitoring newly generated process events in real time based on the cross-departmental collaboration knowledge graph to obtain new process event data units, and matching the new process event data units with nodes in the cross-departmental collaboration knowledge graph to identify collaboration path segments.

Description

Technical Field

[0001] This invention relates to the field of next-generation information technology, and in particular to an enterprise cross-departmental process collaboration management system based on big data resources. Background Technology

[0002] Early enterprise cross-departmental process collaboration management technologies relied heavily on offline manual document transfer and cross-departmental email communication. This resulted in low process flow efficiency and a lack of unified tracking of process status, leading to significant information barriers in cross-departmental collaboration. With the popularization of enterprise information systems, independent business systems such as ERP, CRM, and OA have been widely adopted. Process operations have been recorded online, and process data from each department can be stored and retrieved within their respective systems, thus establishing a preliminary digital foundation for cross-departmental process collaboration.

[0003] Existing enterprise cross-departmental process collaboration management systems mostly focus on monitoring nodes in single-department processes, lacking systematic mining of resource transfer relationships and temporal correlations between cross-departmental process events. They cannot build a complete cross-departmental collaboration knowledge graph, making it difficult for enterprises to clearly grasp the collaboration logic and resource flow paths of cross-departmental processes. Moreover, they are mostly reactive responses after process anomalies occur, unable to proactively extrapolate potential subsequent collaboration paths based on the current process status, and unable to identify high-latency nodes and low-confidence edges in the path in advance. They lack proactive collaboration intervention mechanisms, which can easily lead to problems such as process stagnation and poor resource flow. Therefore, this paper proposes an enterprise cross-departmental process collaboration management system based on big data resources. Summary of the Invention

[0004] In order to overcome the above-mentioned defects of the prior art and to achieve the above objectives, the present invention proposes the following technical solution: A cross-departmental collaborative management system for enterprises based on big data resources includes: Enterprise data acquisition module: Collects raw process log data generated by multiple independent systems within the enterprise, performs unified semantic parsing preprocessing on the raw process log data, and generates cross-departmental process event flows; Data Analysis Module: Based on cross-departmental process event flow, cross-departmental temporal correlation pattern mining is performed to identify effective cross-departmental event sequences. Based on process events in the cross-departmental event sequences as nodes and resource transfer relationships between events as edges, a cross-departmental collaborative knowledge graph is constructed. Decision Collaboration Module: Based on the cross-departmental collaborative knowledge graph, it monitors newly generated process events in real time to obtain new process event data units, and matches the new process event data units with nodes in the cross-departmental collaborative knowledge graph to identify collaborative path segments; Collaborative Management Module: Performs potential collaborative evolution reasoning on collaborative path segments to obtain potential collaborative evolution paths, identifies high-latency nodes and low-confidence edges in the predicted paths of potential collaborative evolution paths to construct an abnormal instruction set, matches the instruction types of the abnormal instruction set with preset collaborative instruction rules, and drives the target department to perform collaborative operations.

[0005] The process of generating cross-departmental process event flows includes: For each independent business system within the enterprise, the original process log data A is obtained. The operation type field in the original process log data A is standardized and mapped to obtain log records with unified semantic parsing. Metadata completion and anomaly filtering are performed on each log record with unified semantic parsing to remove records with missing key fields or timestamp conflicts, forming a standardized process event sequence. ; in, Indicates the first One implementing entity, Indicates the first Each operation type Indicates the first One resource object, Indicates the first A standard timestamp; All standardized process event sequences According to standard timestamps Arranged in ascending order, forming a cross-departmental process event flow with consistent timing. .

[0006] The process of identifying a valid cross-departmental event sequence includes: Based on the process ID, the cross-departmental process event flow B is decomposed into several independent process event sub-sequences; use The sequence pattern mining algorithm extracts valid cross-departmental event subsequences and removes invalid sequences from the valid cross-departmental event subsequences s that conflict with the company's actual business rules, thus obtaining the valid cross-departmental event sequence. ; in, This represents the support of a valid cross-departmental event subsequence s in the cross-departmental process event flow B. This is the preset minimum support threshold.

[0007] The process of building a cross-departmental collaborative knowledge graph includes: Traverse all valid cross-departmental event subsequences in the valid cross-departmental event sequence S, and extract each process event contained in the sequence as a unique node of the knowledge graph; Each process event is mapped one-to-one with the cross-departmental process event flow B, and all extracted and mapped nodes are integrated to form a node set V of the cross-departmental collaborative knowledge graph. Traverse the valid cross-departmental event sequence S, identify the resource flow direction of adjacent process events in the sequence as an edge, and take the process event node that flows out of the resource as the starting node of the edge. The process event node of resource inflow serves as the edge endpoint node. ; Each edge is assigned an attribute, including resource type and resource flow time interval. All edges that have been identified and defined with attributes are integrated to form the edge set E of the cross-departmental collaborative knowledge graph.

[0008] The process of obtaining new process event data units includes: Using a cross-departmental collaborative knowledge graph G as the monitoring benchmark, new process events generated by various independent systems within the enterprise are captured in real time. By establishing a long-connection listening channel through the real-time log interfaces of various enterprise business systems, semantic parsing is performed on the new raw process log data detected, generating new process event data units. .

[0009] The process of identifying cooperative path segments includes: With a new process event data unit for real-time monitoring Using a cross-departmental collaborative knowledge graph G as input, and new process event data units... The core matching dimensions are the execution subject o, operation type h, and resource object r, and attribute consistency is compared with the node v in the cross-departmental collaborative knowledge graph G. Traverse the knowledge graph node set V and filter out the data units related to the new process event. Matching nodes with exact attribute matches To obtain the matching nodes Starting from the cross-departmental collaborative knowledge graph G, traverse the edge set E and extract the edges to match the nodes. All subsequent nodes connected by edges originating from a given point form a collaborative path segment P corresponding to the new event.

[0010] The process of obtaining potential co-evolutionary paths includes: Starting with the terminal node of the collaborative path segment P, and based on the cross-departmental collaborative knowledge graph G, all edge relationships of that node are extracted. Then, based on the resource transfer direction and temporal interval of the edges, the initial collaborative evolution path of the collaborative path segment P is generated by extending along the edge relationships of the cross-departmental collaborative knowledge graph G. ; Traversing potential co-evolutionary paths For each path in the process, perform sequential logic checks and business rule checks, and remove invalid paths that fail the checks from the initial collaborative evolution path. By eliminating the sub-paths, potential co-evolutionary paths are obtained. .

[0011] The process of constructing an exception instruction set includes: Based on potential co-evolutionary paths For each node in the cross-departmental collaborative knowledge graph G, retrieve the historical execution time data of that node, calculate the standard average execution time of the node, and if the predicted execution time of the node exceeds the product of the standard average execution time and the preset delay coefficient, then the node is determined to be a high-latency node. Targeting potential co-evolutionary paths For each edge in the cross-departmental collaborative knowledge graph G, retrieve the confidence attribute of that edge. ; Here, 'e' represents an edge in the cross-departmental collaborative knowledge graph G. This represents the total number of cross-departmental event sequences that include this edge. Starting node of the edge The total number of all cross-departmental event sequences at the beginning; The confidence attribute Conf is lower than the preset confidence threshold. The edges are classified as low-confidence edges; The identified high-latency nodes and low-confidence edges are uniformly marked and summarized to form an abnormal instruction set M.

[0012] The present invention has the following beneficial effects: 1. Based on cross-departmental process event flow, we perform temporal correlation pattern mining to identify cross-departmental event sequences. We then construct a cross-departmental collaborative knowledge graph with process events as nodes and resource transfer relationships as edges. This explicitly represents the temporal correlation relationships and resource flow logic in cross-departmental processes, enabling the visualization and structured expression of cross-departmental collaborative relationships within the enterprise. This provides model support for the analysis and optimization of process collaboration.

[0013] 2. It enables forward-looking prediction and proactive intervention in cross-departmental process collaboration. Based on the cross-departmental collaborative knowledge graph, it monitors new process events in real time, matches graph nodes and identifies collaborative path segments, and performs deduction and compliance verification on subsequent potential collaborative evolution links. It identifies high-latency nodes and low-confidence edges in the path in advance and automatically generates proactive collaboration instructions, realizing the transformation from "passive response after anomalies" to "proactive intervention before risks", effectively reducing the risk of process delays and collaboration failures, and improving the overall flow efficiency of cross-departmental processes.

[0014] 3. Improved the accuracy and effectiveness of cross-departmental collaborative instructions. The proactive collaborative instructions are generated based on the node attributes, edge confidence and real-time process status of the knowledge graph. They contain core information such as the target department, target system, collaborative operation type and execution time. They have clear targeting and executability and can directly drive the target department or system to execute the corresponding process collaborative operation, which solves the problems of insufficient targeting and poor execution effect of traditional collaborative instructions. Attached Figure Description

[0015] Figure 1 This is a system block diagram of the enterprise cross-departmental process collaboration management system based on big data resources proposed in this invention. Detailed Implementation

[0016] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0017] Example 1: As Figure 1 As shown, the enterprise cross-departmental process collaboration management system based on big data resources proposed in this invention includes: Enterprise data acquisition module: Collects raw process log data generated by multiple independent systems within the enterprise, performs unified semantic parsing preprocessing on the raw process log data, and generates cross-departmental process event flows; Collect all original process log data from various independent business systems within the enterprise, including ERP, CRM, OA, and production management systems. The data collection process covers the entire lifecycle of each system's workflow operation records, including full-dimensional log information such as workflow initiation, node approval, resource transfer, and status changes, ensuring the integrity and comprehensiveness of the collected data and obtaining the original workflow log data A. Furthermore, the original process log data A contains heterogeneous structures, non-uniform fields, multi-source timestamps, and fuzzy semantic descriptions, which cannot be directly used for cross-departmental collaborative analysis. In order to achieve cross-system semantic alignment and structured expression, it is necessary to perform unified semantic parsing and attribute alignment preprocessing operations on the original process log data A. The preprocessing steps for unified semantic parsing and attribute alignment of raw process log data include: The operation type field in the original process log data A is standardized and mapped. Based on a predefined ontology library, the operation process described in natural language is converted into a unified semantic tag set. At the same time, the execution subject field is uniformly mapped to the organization coding system, the resource object field is mapped to the asset coding system, and the timestamp field is uniformly converted to... Standard time format for log records that achieve unified semantic parsing; Subsequently, metadata completion and anomaly filtering are performed on each log record with unified semantic parsing. Records with missing key fields or timestamp conflicts are removed, while retaining the complete structure, namely subject-operation-object-time, forming a standardized process event sequence, with each element... Defined as a quadruple: ; Furthermore, Indicates the first One implementing entity, Indicates the first Each operation type Indicates the first One resource object, Indicates the first A standard timestamp; All standardized process event sequences According to standard timestamps Arranged in ascending order, forming a cross-departmental process event flow with consistent timing. , For the first The cross-departmental process event flow corresponding to each standard timestamp is used for subsequent temporal association pattern mining and knowledge graph construction.

[0018] Data Analysis Module: Based on cross-departmental process event flow, cross-departmental temporal correlation pattern mining is performed to identify effective cross-departmental event sequences. Based on process events in the cross-departmental event sequences as nodes and resource transfer relationships between events as edges, a cross-departmental collaborative knowledge graph is constructed. First, cross-departmental temporal correlation pattern mining is performed based on cross-departmental process event flow. The process includes: Based on the process ID, the cross-department process event flow B is decomposed into several independent process event sub-sequences. Each sub-sequence corresponds to a complete cross-department business process, ensuring that the events within the sub-sequence strictly follow the chronological order. use The sequence pattern mining algorithm mines the subsequences of each process event. The mining process takes the temporal relationship between events as the core constraint and extracts frequent subsequences that meet the minimum support threshold. Specifically, these frequent subsequences are effective cross-departmental event subsequences s, reflecting the high-frequency and stable collaborative operation logic in cross-departmental processes of an enterprise.

[0019] All valid cross-departmental event subsequences s obtained from the mining are subjected to business rationality verification, and invalid sequences that conflict with the company's actual business rules are removed, ultimately yielding the valid cross-departmental event sequences. ; in, This represents the support of a valid cross-departmental event subsequence s in the cross-departmental process event flow B. This is the preset minimum support threshold; Based on the obtained effective cross-departmental event sequence S, a node system for the knowledge graph is constructed: First, node extraction is performed by traversing all valid cross-department event subsequences in the valid cross-department event sequence S and extracting each process event contained in the sequence as a unique node of the knowledge graph. Each process event is mapped one-to-one with the cross-departmental process event flow B to ensure that the node attributes fully inherit the core characteristics of the process event, including the executing subject o, operation type h, resource object r, and standard timestamp t. Integrate all extracted and mapped nodes to form a node set V of a cross-departmental collaborative knowledge graph: ; Where v represents a knowledge graph node. For the first A cross-departmental process event flow corresponding to a standard timestamp, where s is a valid cross-departmental event subsequence; Using an effective cross-departmental event sequence S as the core, construct the edge system of a cross-departmental collaborative knowledge graph: Traverse the valid cross-departmental event sequence S, and identify the resource flow direction of adjacent process events in the sequence as an edge, that is: Use the process event node of resource outflow as the edge starting node. The process event node of resource inflow serves as the edge endpoint node. ; Then, edge attributes are defined, and attributes are assigned to each edge. The attributes include resource type, resource transfer time interval, and confidence attribute is assigned to the edge based on the frequency of occurrence of events in the sequence, which represents the business reliability of the resource transfer relationship. Next, the edge set is generated, integrating all edges after identification and attribute definition to form the edge set E of the cross-departmental collaborative knowledge graph: ; Where e represents a knowledge graph edge. Let it be the starting node of the edge. Let it be the endpoint node of the edge. The confidence attribute of an edge. For the time interval of resource flow, the edge set E is generated entirely by the effective cross-departmental event sequence S and the node set V of the cross-departmental collaborative knowledge graph; Finally, by integrating the node set V and the edge set E, a complete cross-departmental collaborative knowledge graph is generated. .

[0020] Decision Collaboration Module: Based on the cross-departmental collaborative knowledge graph, it monitors newly generated process events in real time to obtain new process event data units, and matches the new process event data units with nodes in the cross-departmental collaborative knowledge graph to identify collaborative path segments; Using a cross-departmental collaborative knowledge graph G as the monitoring benchmark, new process events generated by various independent systems within the enterprise are captured in real time. By establishing long-connection listening channels through the real-time log interfaces of various business systems of the enterprise, covering all independent systems such as ERP, CRM, and OA, we can ensure that no newly generated process operation logs are missed. For the new raw process log data detected, perform semantic parsing to generate new process event data units. It has unified semantic and attribute specifications with cross-departmental process event flow B; All standardized new process event data units They are added to the real-time listening queue in chronological order of their generation, providing an ordered data source for subsequent matching operations; Perform matching of new events with knowledge graph nodes: With a new process event data unit for real-time monitoring With the cross-departmental collaborative knowledge graph G as input, perform matching operations between new process events and graph nodes to identify the corresponding collaborative path segments; With new process event data units The core matching dimensions are the execution subject o, operation type h, and resource object r, and attribute consistency is compared with the node v in the cross-departmental collaborative knowledge graph G. Then, node matching is performed, traversing the knowledge graph node set V to filter out data units that match the new process event. Matching nodes with exact attribute matches This node is the corresponding entity of the new process event in the graph; Simultaneously, collaborative path segment identification is performed to match the obtained matching nodes. Starting from the cross-departmental collaborative knowledge graph G, traverse the edge set E and extract the edges to match the nodes. All subsequent nodes connected by edges originating from a given point form a collaborative path segment corresponding to the new event. ; Where k is the node index and P is the cooperative path segment. For matching nodes in the graph Subsequent nodes with edge relationships.

[0021] Collaborative Management Module: Performs potential collaborative evolution reasoning on collaborative path segments to obtain potential collaborative evolution paths, identifies high-latency nodes and low-confidence edges in the predicted paths of potential collaborative evolution paths to construct an abnormal instruction set, matches the instruction types of the abnormal instruction set with preset collaborative instruction rules, and drives the target department to perform collaborative operations. Reasoning about the potential co-evolutionary paths of co-evolutionary path segments: Taking the terminal node of the collaborative path segment P as the starting point of reasoning, based on the cross-departmental collaborative knowledge graph G, all outgoing edge relationships of the node are extracted, and all subsequent process event nodes that can be associated with the starting node are identified, providing basic constraints for path deduction. Based on the enterprise's pre-defined business process rules, the execution sequence and logical deduction are performed for each outgoing edge relationship: Based on the resource transfer direction and temporal interval of the edges, the initial collaborative evolution path of the collaborative path segment P is generated by extending layer by layer along the edge relationships of the cross-departmental collaborative knowledge graph G. ; Each path within the set corresponds to a potential cross-departmental collaborative process following the new event, and all nodes and edges of the paths come from the cross-departmental collaborative knowledge graph G. Traversing the initial co-evolution path Perform double verification on each path in the process: Timing logic verification: Verify whether the timestamp order of all nodes in the path meets the process timing requirements, and remove invalid paths with time reversal or timing conflicts; Business rule verification: Verify whether the departmental flow, resource transfer, and operation type of the path comply with the enterprise's business specifications, and remove invalid paths that do not comply with process permissions and resource constraints; Invalid paths that fail the verification will be removed from the initial co-evolutionary path. By eliminating the sub-paths and retaining only those that simultaneously satisfy both temporal logic and business rules, we obtain the final potential collaborative evolution path. ; Then, high-latency nodes and low-confidence edges are identified in the potential co-evolutionary path prediction paths: The high-latency node identification process includes: Based on potential co-evolutionary paths For each node in the cross-departmental collaborative knowledge graph G, retrieve the historical execution time data of that node and calculate the standard average execution time of the node; If the predicted execution time of a node exceeds the product of the standard average execution time and the preset latency coefficient, the node is determined to be a high-latency node. The calculation process for the standard average execution time is as follows: ; in, The standard average execution time for node v. Let be the execution duration of the i-th historical execution of node v. This refers to the historical number of executions. If the standard average execution time of the node Exceeding the preset execution time If so, then this node is determined to be a high-latency node; The low-confidence edge identification process includes: Targeting potential co-evolutionary paths For each edge in the cross-departmental collaborative knowledge graph G, retrieve the confidence attribute of that edge. ; Here, 'e' represents an edge in the cross-departmental collaborative knowledge graph G, corresponding to the node starting from the edge. to the edge endpoint node Resource transfer relationships, This represents the total number of cross-departmental event sequences that include this edge. Starting node of the edge The total number of all cross-departmental event sequences at the beginning; Specifically, confidence attribute The higher the confidence level, the more stable the resource transfer relationship is in the company's historical processes and the more it conforms to the normal business logic. The lower the confidence level, the lower the frequency of the relationship and the weaker the reliability of collaboration, and the greater the risk of process interruption and collaboration failure. If the confidence attribute Conf of an edge is lower than the preset confidence threshold If the edge is low confidence, it is considered a low confidence edge, indicating that the reliability of the resource transfer relationship is insufficient. The identified high-latency nodes and low-confidence edges are uniformly marked and summarized to form an abnormal instruction set M; Traverse the abnormal instruction set M, extract the department, operation type, and resource requirements of high-latency nodes, as well as the transmission resources and target node parameters of low-confidence edges, as the basic content of proactive collaboration instructions; Based on the instruction type of the abnormal instruction set M, a preset cooperative instruction rule is matched, including: For high-latency nodes, match expedited processing and resource priority allocation command rules; For low-confidence edges, match process intervention and collaborative reminder rules; Based on the matching instruction rules and the extracted core parameters, a standardized proactive collaborative instruction Q is generated; The proactive collaboration instruction Q includes core information such as the target department identifier, target system identifier, collaboration operation type, and execution time requirements, ensuring the executability and relevance of the instruction; Then, the proactive collaboration instructions are executed to drive the target departments to perform collaborative operations. Each instruction in the proactive collaboration instructions is distributed to the corresponding target execution entity's system or user terminal to trigger collaborative actions. After the actions are executed, the system records the execution results, including the action completion time, execution status, and resource status changes, forming an execution log to complete cross-departmental process collaboration management.

[0022] In the application, several formulas are calculated by removing dimensions and taking their numerical values. The formulas are established by collecting a large amount of data and simulating the most recent real situation. Some coefficients or weights in the formulas are set by those skilled in the art according to the actual situation, so they will not be elaborated here.

[0023] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented in software, the above embodiments can be implemented, in whole or in part, as a computer program product. Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution.

[0024] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. An enterprise cross-departmental process collaboration management system based on big data resources, characterized in that: include: Enterprise data acquisition module: Collects raw process log data generated by multiple independent systems within the enterprise, performs unified semantic parsing preprocessing on the raw process log data, and generates cross-departmental process event flows; Data Analysis Module: Based on cross-departmental process event flow, cross-departmental temporal correlation pattern mining is performed to identify effective cross-departmental event sequences. Based on process events in the cross-departmental event sequences as nodes and resource transfer relationships between events as edges, a cross-departmental collaborative knowledge graph is constructed. Decision Collaboration Module: Based on the cross-departmental collaborative knowledge graph, it monitors newly generated process events in real time to obtain new process event data units, and matches the new process event data units with nodes in the cross-departmental collaborative knowledge graph to identify collaborative path segments; Collaborative Management Module: Performs potential collaborative evolution reasoning on collaborative path segments to obtain potential collaborative evolution paths, identifies high-latency nodes and low-confidence edges in the predicted paths of potential collaborative evolution paths to construct an abnormal instruction set, matches the instruction types of the abnormal instruction set with preset collaborative instruction rules, and drives the target department to perform collaborative operations.

2. The enterprise cross-departmental process collaboration management system based on big data resources according to claim 1, characterized in that, The process of generating cross-departmental process event flows includes: For each independent business system within the enterprise, the original process log data A is obtained. The operation type field in the original process log data A is standardized and mapped to obtain log records with unified semantic parsing. Metadata completion and anomaly filtering are performed on each log record with unified semantic parsing to remove records with missing key fields or timestamp conflicts, forming a standardized process event sequence. ; in, Indicates the first One implementing entity, Indicates the first Each operation type Indicates the first One resource object, Indicates the first A standard timestamp; All standardized process event sequences According to standard timestamps Arranged in ascending order, forming a cross-departmental process event flow with consistent timing. .

3. The enterprise cross-departmental process collaboration management system based on big data resources according to claim 2, characterized in that, The process of identifying a valid cross-departmental event sequence includes: Based on the process ID, the cross-departmental process event flow B is decomposed into several independent process event sub-sequences; use The sequence pattern mining algorithm extracts valid cross-departmental event subsequences and removes invalid sequences from the valid cross-departmental event subsequences s that conflict with the company's actual business rules, thus obtaining the valid cross-departmental event sequence. ; in, This represents the support of a valid cross-departmental event subsequence s in the cross-departmental process event flow B. This is the preset minimum support threshold.

4. The enterprise cross-departmental process collaboration management system based on big data resources according to claim 3, characterized in that, The process of building a cross-departmental collaborative knowledge graph includes: Traverse all valid cross-departmental event subsequences in the valid cross-departmental event sequence S, and extract each process event contained in the sequence as a unique node of the knowledge graph; Each process event is mapped one-to-one with the cross-departmental process event flow B, and all extracted and mapped nodes are integrated to form a node set V of the cross-departmental collaborative knowledge graph. Traverse the valid cross-departmental event sequence S, identify the resource flow direction of adjacent process events in the sequence as an edge, and take the process event node that flows out of the resource as the starting node of the edge. The process event node of resource inflow serves as the edge endpoint node. ; Each edge is assigned an attribute, including resource type and resource flow time interval. All edges that have been identified and defined with attributes are integrated to form the edge set E of the cross-departmental collaborative knowledge graph.

5. The enterprise cross-departmental process collaboration management system based on big data resources according to claim 4, characterized in that, The process of obtaining new process event data units includes: Using a cross-departmental collaborative knowledge graph G as the monitoring benchmark, new process events generated by various independent systems within the enterprise are captured in real time. By establishing a long-connection listening channel through the real-time log interfaces of various enterprise business systems, semantic parsing is performed on the new raw process log data detected, generating new process event data units. .

6. The enterprise cross-departmental process collaboration management system based on big data resources according to claim 5, characterized in that, The process of identifying cooperative path segments includes: With a new process event data unit for real-time monitoring Using a cross-departmental collaborative knowledge graph G as input, and new process event data units... The core matching dimensions are the execution subject o, operation type h, and resource object r, and attribute consistency is compared with the node v in the cross-departmental collaborative knowledge graph G. Traverse the knowledge graph node set V and filter out the data units related to the new process event. Matching nodes with exact attribute matches To match nodes Starting from the cross-departmental collaborative knowledge graph G, traverse the edge set E and extract the edges to match the nodes. All subsequent nodes connected by edges originating from a given point form a collaborative path segment P corresponding to the new event.

7. The enterprise cross-departmental process collaboration management system based on big data resources according to claim 6, characterized in that, The process of obtaining potential co-evolutionary paths includes: Starting with the terminal node of the collaborative path segment P, and based on the cross-departmental collaborative knowledge graph G, all edge relationships of that node are extracted. Then, based on the resource transfer direction and temporal interval of the edges, the initial collaborative evolution path of the collaborative path segment P is generated by extending along the edge relationships of the cross-departmental collaborative knowledge graph G. ; Traversing potential co-evolutionary paths For each path in the process, perform sequential logic checks and business rule checks, and remove invalid paths that fail the checks from the initial collaborative evolution path. By eliminating the sub-paths, potential co-evolutionary paths are obtained. .

8. The enterprise cross-departmental process collaboration management system based on big data resources according to claim 7, characterized in that, The process of constructing an exception instruction set includes: Based on potential co-evolutionary paths For each node in the cross-departmental collaborative knowledge graph G, retrieve the historical execution time data of that node, calculate the standard average execution time of the node, and if the predicted execution time of the node exceeds the product of the standard average execution time and the preset delay coefficient, then the node is determined to be a high-latency node. Targeting potential co-evolutionary paths For each edge in the cross-departmental collaborative knowledge graph G, retrieve the confidence attribute of that edge. ; Here, 'e' represents an edge in the cross-departmental collaborative knowledge graph G. This represents the total number of cross-departmental event sequences that include this edge. Starting node of the edge The total number of all cross-departmental event sequences at the beginning; The confidence attribute Conf is lower than the preset confidence threshold. The edges are classified as low-confidence edges; The identified high-latency nodes and low-confidence edges are uniformly marked and summarized to form an abnormal instruction set M.

Images