Employee behavior evaluation method based on function engine and CRM system

Abstract

The invention relates to the technical field of enterprise data management, in particular to an employee behavior evaluation method based on a function engine and a CRM system.The method comprises the steps that after employee behavior configuration of a business management end is responded, an employee behavior rule is defined, then a process engine is called, behavior data and context information of an employee are packaged into fact data, and the fact data are stored in a database; and the fact data is transmitted to a rule engine through a standardized interface of the CRM system, and finally the rule engine is called to carry out matching analysis operation on the fact data and employee behavior rules to obtain employee behavior scores. Compared with the prior art, the rule engine designed by the invention can carry out matching analysis calculation on the fact data and the employee behavior rule, can adapt to instant modification of various rule definitions and carry out analysis matching, so that various complex employee behaviors are scored, and the problem of poor flexibility in the prior art is solved.

Description

Technical Field

[0001] The present invention relates to the technical field of enterprise data management, and more specifically, to an employee behavior evaluation method and a CRM system based on a function engine. Background Art

[0002] Customer Relationship Management (CRM) systems are essential tools for modern business operations. They not only help companies manage customer interactions and record customer data, but also host a vast amount of data on employees' daily work behaviors. This behavioral data, such as customer visits, sales opportunity follow-ups, and service ticket processing, directly reflects employee performance and their contribution to business goals. Therefore, effectively leveraging CRM system data to scientifically and rationally assess employee behavior is crucial for improving employee performance, optimizing business processes, and achieving corporate strategic goals.

[0003] To address the aforementioned issues, existing technologies, such as Microsoft Dynamics 365, include built-in business rules or workflow features, allowing administrators to customize business logic and automated processes to a certain extent to calculate or analyze employee behavior scores or evaluations. For example, rules can be configured to automatically update fields, display prompts, or lock fields under specific conditions. However, while the Microsoft Dynamics 365 system provides these business rule features, it lacks sufficient flexibility in terms of granular rule versioning, complex event processing capabilities, or the ability for non-IT professionals to intuitively define and manage evaluation rules with complex scoring logic and interdependencies, often requiring customized development by IT professionals.

[0004] Therefore, the existing CRM system has the problem of poor flexibility in analyzing employee behavior. Summary of the Invention

[0005] In order to solve the above technical problem of poor flexibility, the present invention discloses an employee behavior evaluation method and a CRM system based on a function engine.

[0006] In a first aspect, the present invention discloses a method for evaluating employee behavior based on a function engine, which is used in a CRM system integrating a process engine and a rule engine. The method of the present invention comprises: Define employee behavior rules in response to employee behavior configuration on the business management side; Call the process engine to encapsulate employee behavior data and contextual information into fact data, and then transmit the fact data to the rule engine through the standardized interface of the CRM system; The rule engine is called to match and analyze the factual data with the employee behavior rules to obtain the employee behavior score.

[0007] Beneficial effects: Through the above technical solution, after non-IT professionals define employee behavior rules (which can be standardized standard form descriptions or fragmentary colloquial definitions), the process engine that is highly integrated with the CRM system will automatically encapsulate the employee behavior data into fact data that can be read / recognized by the rule engine, and transmit the fact data to the rule engine through a standardized interface, so as to overcome the problem in the existing technology that setting new rules requires IT personnel to perform underlying definitions. On this basis, the rule engine designed by the present invention can match, analyze and calculate fact data with employee behavior rules, can adapt to the instant modification of various rule definitions, and perform analysis and matching, thereby realizing the scoring of various complex employee behaviors and solving the problem of poor flexibility of the existing technology.

[0008] Preferably, the process engine is called to encapsulate the employee's behavior data and context information into fact data, specifically: In response to the generation of employee behavior data in the CRM system, CRM events are captured through the CRM system's API interface or message queue; According to the CRM event type, find and load the corresponding process engine; Import CRM events into the process engine to obtain process instances; In the process instance, extract employee behavior data and context information; Encapsulate behavioral data and contextual information into factual data.

[0009] Beneficial effects: The method of the present invention can automatically search and load the corresponding process engine according to the CRM event type, realize the optimization and self-adaptation of the process engine, so as to achieve better employee behavior data extraction effect, further improving the flexibility of the method of the present invention.

[0010] Preferably, the rule engine adopts a rule engine based on the Rete algorithm or the RetePlus algorithm.

[0011] Preferably, the rule engine is called to match the fact data with the employee behavior rules for analysis and calculation to obtain the employee behavior score, specifically: In response to the generation of fact data, employee behavior rules are loaded into the Rete algorithm or RetePlus algorithm and compiled into a directed acyclic graph matching network; The factual data is input into the directed acyclic graph matching network for data screening, data matching, conflicting data screening and scoring to obtain the employee behavior score.

[0012] Beneficial Effects: The rule engine designed in this invention compiles factual data into an acyclic graph matching network for deeper data processing. Specifically, the data screening phase enables data diversion and the removal of irrelevant data, the data matching phase facilitates data location and assists in scoring operations, and conflicting data screening removes duplicate scoring data. Compared to existing technologies, employee behavior scores generated based on this solution are more accurate, the calculation process is shorter, and it is more adaptable to complex application scenarios, further enhancing the flexibility of the present method.

[0013] Preferably, the directed acyclic graph matching network includes a root node, an object type screening node, an Alpha matching node, a Beta connection node and a terminal node.

[0014] Preferably, the factual data is input into a directed acyclic graph matching network for data screening, specifically: Using the root node as the network entry, fact data is distributed to multiple object type screening nodes; Distribute the object types that meet the preset screening rules in the object type screening node to the corresponding Alpha matching node.

[0015] Preferably, the fact data is input into a directed acyclic graph matching network for data matching, specifically: Use the Alpha matching node as the attribute matching node for a single fact data, and store the fact data that meets the employee behavior rules in the Alpha memory; The Beta connection node is used as an attribute matching node that spans multiple fact data matches. Multiple fact data stored in the Alpha memory or the output data of other Beta matching nodes are connected to obtain matching results.

[0016] Preferably, after obtaining the matching result, the method of the present invention further comprises: The matching results and / or any factual data of completing the evaluation of any employee behavior rule are delivered to the corresponding terminal node.

[0017] Preferably, if multiple terminal nodes are activated at the same time, the factual data is input into the directed acyclic graph matching network to filter out conflicting data and score, and obtain the employee behavior score, specifically: The matching results and / or fact data output by multiple terminal nodes are placed into an agenda queue for sorting, and duplicate matching results or duplicate fact data obtained based on the same employee behavior rule are filtered out from the agenda queue; According to the order of the agenda queue, the matching results and / or factual data are scored item by item to obtain the employee behavior score.

[0018] In a second aspect, the present invention discloses a CRM system for implementing the employee behavior evaluation method based on the function engine described in the first aspect. The system of the present invention comprises at least a dynamic configuration management module, a process engine module, and a rule engine module; wherein, Dynamic configuration management module, used to respond to employee behavior configuration on the business management side and define employee behavior rules; The process engine module is used to encapsulate employee behavior data and contextual information into fact data and transmit the fact data to the rule engine module through the standardized interface of the CRM system; The rule engine module is used to match and analyze factual data with employee behavior rules to obtain employee behavior scores.

[0019] Beneficial effects: The present invention designs a CRM system with a highly integrated process engine and rule engine. The system can automatically respond to the employee behavior rule setting of non-IT professionals, quickly change the evaluation criteria of employee behavior rules, adapt to various emergencies, and has extremely high flexibility.

[0020] The beneficial effects of the present invention are: (1) Compared with the existing technology, the rule engine designed in the present invention can match and analyze fact data with employee behavior rules, and can adapt to the instant modification of various rule definitions and perform analysis and matching, thereby achieving the scoring of various complex employee behaviors and solving the problem of poor flexibility of the existing technology.

[0021] (2) Compared with the existing technology, the employee behavior score calculated based on the rule engine is more accurate, the calculation process takes less time, and it is more adaptable to complex application scenarios, which further enhances the flexibility of the method of the present invention.

[0022] (3) Compared with the existing technology, the method of the present invention can automatically search and load the corresponding process engine according to the CRM event type, realize the optimization and self-adaptation of the process engine, so as to achieve better employee behavior data extraction effect, further improving the flexibility of the method of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 This is a flow chart of the employee behavior evaluation method based on the function engine in the first embodiment of the present invention; Figure 2 This is an architecture diagram of the CRM system in the second embodiment of the present invention. DETAILED DESCRIPTION

[0024] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, but not all of the embodiments.

[0025] The specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0026] like Figure 1 As shown, this embodiment discloses an employee behavior evaluation method based on a function engine, which is used in a CRM system integrating a process engine and a rule engine.

[0027] Specifically, the method of this embodiment includes: S10: In response to the employee behavior configuration on the business management side, define employee behavior rules.

[0028] In this embodiment, authorized users (business management) of the CRM system are allowed to dynamically define employee behavior types (e.g., visiting a customer, creating a sales opportunity, or handling a service ticket) through a graphical user interface or configuration file, and configure corresponding evaluation rules for each behavior type. These rules include trigger conditions (LHS, Left Hand Side) and execution actions (RHS, Right Hand Side, primarily the logic for calculating points). The definition or description of employee behavior rules can be standardized, tabular, or fragmented, and can be more colloquial.

[0029] S20: Call the process engine, encapsulate the employee's behavior data and context information into fact data, and transmit the fact data to the rule engine through the standardized interface of the CRM system.

[0030] In this embodiment, the process engine is a rule engine based on the Rete algorithm or the RetePlus algorithm. The standardized interface can be a standardized interface provided by the CRM system, such as a RESTful API, a message queue subscription, or a database view.

[0031] S30: Call the rule engine to match and analyze the fact data with the employee behavior rules to obtain the employee behavior score.

[0032] Through the above steps S10 to S30, the method of the present invention can define employee behavior rules for non-IT professionals. The process engine that is highly integrated with the CRM system will automatically encapsulate employee behavior data into fact data that can be read / recognized by the rule engine, and transmit the fact data to the rule engine through a standardized interface, so as to overcome the problem that setting new rules in the prior art requires IT personnel to perform underlying definitions. On this basis, the rule engine designed by the present invention can match, analyze and calculate fact data with employee behavior rules, can adapt to the instant modification of various rule definitions, and perform analysis and matching, thereby realizing the scoring of various complex employee behaviors and solving the problem of poor flexibility of the prior art.

[0033] Furthermore, the employee behavior configuration in step S10 includes employee behavior definition, rule definition (trigger conditions, trigger actions, and configured points), and weight and threshold configuration. This employee behavior configuration can be implemented through the dynamic configuration management module in the CRM system. Business management can flexibly define and adjust evaluation criteria based on enterprise management needs and business priorities.

[0034] Specifically, based on the above employee behavior definitions, the business management side first defines the employee behavior types that need to be tracked and evaluated. These behavior types are directly derived from employees' daily operations in the CRM system.

[0035] For example, the employee behavior types mentioned above can be defined as "adding a new valid customer visit", "successfully closing a high-value business opportunity", "completing a service ticket on time", "initiating a team knowledge sharing" and / or "successfully upselling a product on a new media channel". Each behavior type will be assigned a unique identifier and can be associated with one or more specific data objects in the CRM system (such as a "visit record" object, a "business opportunity" object, a "work order" object) or a specific event (such as "business opportunity stage changed to won order", "work order status changed to resolved"). Among them, the behavior list provided by the business management end (such as visitor information, visit behavior, new business development points, etc.) serves as an important reference for behavior definition.

[0036] Specifically, for the above rule definitions, the business management side needs to configure one or more specific evaluation rules. These rules are usually configured in an easy-to-understand "if-then" logical structure or in a visual form such as a decision table or decision tree.

[0037] The condition part of the rule definition defines the specific conditions that must be met for the behavior to be considered valid and trigger the calculation of points. For example: For the "Add a New Valid Customer Visit" action: IF the customer level = KA key account AND the visit duration (minutes) >= 30 AND the visit report completeness > 80%, then trigger the corresponding action process. For the "Successfully Close a High-Value Opportunity" action: IF the opportunity amount (RMB) > 100,000 AND the product line = core product, then trigger the corresponding action process. For the "Complete a Service Ticket on Time" action: IF the ticket priority = 'High' AND the resolution time <= the promised resolution time, then trigger the corresponding action process.

[0038] That is, the trigger of the above conditional part can be various data fields recorded in CRM (text, numbers, dates, option sets, etc.), attributes of the behavior itself (such as number of occurrences, duration), or intermediate values ​​obtained through simple calculations.

[0039] The action part of the rule definition defines the operation that the CRM system should perform when the condition part (LHS) is met. In this embodiment, the core operation is to assign points to the corresponding behavior. The scoring strategy can be: (1) Fixed scoring: A fixed score is given if the conditions are met.

[0040] (2) Dynamic scoring: The score is dynamically calculated based on one or more variables in the behavioral data, such as calculating commission points based on a certain percentage of the sales amount.

[0041] (3) Lookup and assign points: Look up the corresponding score in a preset point comparison table based on a certain attribute of the behavioral data (such as product category, customer level).

[0042] (4) Tiered scoring: Different points are awarded based on the different levels of behavioral performance (such as different ranges of completion rate).

[0043] Regarding weighting and threshold configuration, to reflect the varying contributions of different behaviors to business goals, business management can set weighting coefficients for different behavior categories (such as "Customer Development," "Customer Service," and "Team Collaboration") or for individual specific behaviors. An employee's final total score may be the weighted sum of the scores for each behavior. When an employee's score for a specific category or their total score reaches a preset threshold, they can earn additional bonus points, special badges, or even be promoted to a higher rating level.

[0044] After executing step S10, the employee behavior configuration (including behavior definitions, rule conditions, score values, weights, and thresholds) is stored in the CRM system's data storage module. This can be modified, added, deleted, enabled, or disabled online at any time through the CRM system's dynamic configuration management module. These adjustments typically take effect quickly without requiring a CRM system restart or underlying code changes. Table 1 below shows the CRM system rules table adaptively generated based on the employee behavior configuration described above: Table 1

[0045] It should be noted that the above Table 1 is only a sample rule configuration display, and does not mean that all employee behavior rule configurations must be configured according to the items in the above table.

[0046] Furthermore, the above step S20 specifically includes: S21: In response to the generation of employee behavior data in the CRM system, a CRM event is captured through the API interface or message queue of the CRM system.

[0047] S22: According to the CRM event type, search and load the corresponding process engine.

[0048] S23: Import the CRM event into the process engine to obtain a process instance.

[0049] S24: In the process instance, extract employee behavior data and context information.

[0050] S25: Encapsulate the behavior data and context information into fact data.

[0051] In this embodiment, multiple or various process engines are provided in the process engine module.

[0052] Through steps S21 to S25, the method of this embodiment can automatically search and load the corresponding process engine according to the CRM event type, realize the optimization and adaptation of the process engine, so as to achieve better employee behavior data extraction effect, further improving the flexibility of the method of the present invention.

[0053] In combination with the above steps S21 to S25, the technical effects are described below with reference to examples: When a salesperson saves a new customer visit record in the CRM system, the CRM system can notify the process engine through API calls, sending messages to message queues (such as Kafka and RabbitMQ), or triggering database triggers. Once the process is triggered, the process engine starts a new process instance. This instance, based on the process definition, executes a series of service tasks to collect complete data related to the action. For example, for the "Add a New Visit Record" action, the process engine collects the actor (employee ID), the time of the action, associated customer information (customer ID, name, level), the type of visit, the duration of the visit, and any completed visit notes. This data constitutes the contextual information needed to evaluate the action. When the process reaches the "Rule Task" node, the process engine uses the constructed contextual data (usually encapsulated as one or more fact objects) as input and calls the rule engine's evaluation interface. Throughout the process execution, the process engine continuously manages and updates the state of the process instance and persistently stores key state information (such as the current execution node or the values ​​of process variables) in the data storage module. This ensures that after a CRM system failure or restart, the ongoing process can be restored from the interruption point, ensuring the reliability and consistency of data processing.

[0054] Furthermore, the above step S30 is specifically as follows: S31: In response to the generation of fact data, the employee behavior rules are loaded into the Rete algorithm or the RetePlus algorithm and compiled into a directed acyclic graph matching network.

[0055] S32: Input the fact data into the directed acyclic graph matching network to perform data screening, data matching, conflicting data screening and scoring to obtain the employee behavior score.

[0056] Through steps S31 and S32, the rule engine disclosed in this embodiment can compile factual data into an acyclic graph matching network for deeper data processing. Data screening is used for data diversion and removal of irrelevant data, data matching is used for data location and auxiliary scoring operations, and conflict data is used to remove duplicate scoring data.

[0057] Compared with the existing technology, the employee behavior score obtained based on the above scheme is more accurate, the calculation process takes less time, and it is more adaptable to complex application scenarios, which further enhances the flexibility of the method of the present invention.

[0058] Furthermore, the directed acyclic graph matching network includes a root node, an object type screening node, an alpha matching node, a beta connection node, and a terminal node. The following is a detailed description of the design architecture of the directed acyclic graph: Root Node: Serves as the entry point for factual data into the network.

[0059] Object Type Node (Alpha Network Entry): Performs preliminary type filtering on fact data. For example, only objects of type VisitFact are allowed to pass.

[0060] Alpha Matching Node (AlphaNode): In the Alpha network, performs intra-element tests on individual fact data. For example, one Alpha node might test that VisitFact.type == 'SKA customer', while another Alpha node tests that VisitFact.duration > 30. Alpha nodes that meet the conditions will pass their matching fact data to their output and may store it in Alpha memory.

[0061] BetaNode: Performs join operations across multiple facts or inter-element tests on fact sets in the Beta network. Beta nodes typically have two inputs, one from Alpha memory or the output of another Beta node. For example, a Beta node might join a VisitFact that satisfies type == 'SKA Customer' with an EmployeeFact that satisfies employeeLevel == 'Senior' (if the rule requires employee level information). The Beta node stores the partial matches of successful joins in its Beta memory.

[0062] Terminal Node (P-Node): When a fact or combination of facts successfully passes through all nodes representing the complete left-hand side (LHS) of a rule in the Rete network, it reaches the corresponding terminal node for that rule. This indicates that the rule is activated. A key advantage of the Rete algorithm is node sharing: if multiple rules share the same conditional part, they share the corresponding node in the Rete network, thus avoiding duplicate computation.

[0063] It's important to note that during the aforementioned fact data transmission process, when new facts are asserted into the working memory, they propagate through the Rete network starting from the root node. Whenever a fact passes through an Alpha node or a group of facts passes through a Beta node, the corresponding memory (Alpha memory or Beta memory) is updated, and the matching results are propagated to downstream nodes.

[0064] Furthermore, in the above step S32, the fact data is input into the directed acyclic graph matching network for data screening, specifically: S321: Using the root node as a network entry, distribute fact data to multiple object type screening nodes.

[0065] S322: Distribute the object types that meet the preset screening rules in the object type screening node to the corresponding Alpha matching node.

[0066] Through the above steps S321 and S322, concurrent data screening can be performed through multiple Alpha matching nodes, which can enhance the data throughput of the CRM system, speed up data transmission, and improve the data response capability of the CRM system.

[0067] Furthermore, in the above step S32, the fact data is input into the directed acyclic graph matching network for data matching, specifically: S323: Using the Alpha matching node as the attribute matching node for a single fact data, the fact data that meets the employee behavior rules is stored in the Alpha memory.

[0068] S324: Using the Beta connection node as an attribute matching node that spans multiple fact data matches, multiple fact data stored in the Alpha memory or output data of other Beta matching nodes are data connected to obtain a matching result.

[0069] Through steps S323 and S324, information matching can be performed concurrently through Alpha matching nodes or Beta connection nodes, further improving the data processing efficiency of the CRM system. Furthermore, by calling multiple Alpha matching nodes and Beta connection nodes, nearly all possible matching results can be generated to address the complexity of employee behavior data, further enhancing the flexibility of the method of this embodiment.

[0070] Furthermore, after obtaining the matching result in step S324, the method of this embodiment further includes: The matching results and / or any factual data of completing the evaluation of any employee behavior rule are delivered to the corresponding terminal node.

[0071] If multiple terminal nodes are activated at the same time, the fact data is input into the directed acyclic graph matching network in step S32 to filter out conflicting data and score the employee behavior score, which is specifically: The matching results and / or fact data output by multiple terminal nodes are placed in an agenda queue for sorting, and duplicate matching results or duplicate fact data obtained based on the same employee behavior rule are filtered out from the agenda queue.

[0072] According to the order of the agenda queue, the matching results and / or factual data are scored item by item to obtain the employee behavior score.

[0073] The sorting of the agenda queue can be achieved by using the following conflict resolution strategy: First, sort by priority: pre-set priorities for the rules, with high-priority rules sorted at the top and sub-item scoring executed first.

[0074] Second, sort by recency or newness: rules with facts that have recently entered the working memory are sorted first, and sub-item scoring is performed first.

[0075] Third, sort by specificity: rules with more specific conditions (i.e., more complex LHS or more constraints) are sorted higher and sub-item scoring is performed first. With this technical solution, if multiple rules share the same conditional part, they will share the corresponding node in the Rete network, avoiding duplicate calculations. Furthermore, based on the conflict resolution strategy, higher-priority matching results and / or fact data can be scored separately.

[0076] Preferably, after the above step S30, the method of this embodiment further includes: S40: Summarize and visualize the employee behavior scores of each employee.

[0077] Specifically, the system automatically accumulates points earned by each employee for different behaviors and summarizes them based on preset periods (such as daily, weekly, or monthly). Based on these points, a ranking list of employee behaviors is generated, and ratings are automatically assigned based on configured score ranges. The evaluation results are ultimately presented to employees and their managers via CRM dashboards, reports, or other formats.

[0078] Through the above step S40, quantitative ranking and rating of employee behaviors can be achieved, which greatly reduces the workload of managers.

[0079] Based on the above technical description, compared with the existing technology, the method of this embodiment has at least the following advantages: 1. Greater flexibility and adaptability: Criteria, rules, and processes for evaluating employee behavior are no longer hard-coded or fixed in the system. Instead, they can be dynamically created, modified, enabled, or disabled through a user-friendly configuration interface. This enables companies to quickly respond to and adjust evaluation systems based on market changes, business strategy adjustments, or shifts in management priorities, significantly improving management flexibility and adaptability.

[0080] 2. Higher evaluation efficiency and real-time performance: The process engine automatically collects employee behavior data in the CRM, and the rules engine performs real-time evaluation and score calculation, significantly reducing the need for manual statistics and intervention. Employees receive near-real-time feedback on their performance, helping to guide and motivate them in a timely manner.

[0081] 3. Reduced maintenance costs and reliance on IT professionals: Since adjustments to evaluation rules and processes are primarily made through configuration rather than code-level modifications, system maintenance complexity and costs are significantly reduced. Business departments can manage the evaluation system more independently, reducing reliance on IT support and shortening the response time.

[0082] 4. Enhanced transparency and fairness in employee ratings: All behavior types, rules, and scoring criteria are clearly defined and visible to relevant personnel (based on permissions), making the entire employee behavior rating process more transparent. Automated scoring based on unified, objective rules helps improve the fairness and consistency of evaluation results and reduce subjective bias.

[0083] 5. Can promote employee behavior optimization: Through quantitative scoring and ranking incentives for key behaviors, employees can be effectively guided to focus on the behavioral direction expected by the company, such as more active customer visits, more efficient handling of customer issues, more proactive team collaboration and knowledge sharing, etc., thereby promoting overall performance improvement.

[0084] Example 2 like Figure 2 As shown, this embodiment discloses a CRM system for implementing the employee behavior evaluation method based on the function engine recorded in embodiment one or two. The system of the present invention includes a dynamic configuration management module, a process engine module, a rule engine module, a CRM platform integration module and a data storage module.

[0085] The dynamic configuration management module responds to employee behavior configurations on the business management side and defines employee behavior rules. The process engine module encapsulates employee behavior data and contextual information into factual data and transmits this data to the rule engine module via the CRM system's standardized interface. The rule engine module then matches and analyzes the factual data against employee behavior rules to generate employee behavior scores.

[0086] More specifically, the process engine module includes a process definer, a process executor, a task manager, and a persistence mechanism.

[0087] The process definer provides graphical or declarative tools, allowing business managers to design and define business processes for processing diverse employee behavior data. These processes typically follow BPMN (Business Process Model and Notation) or similar specifications, clearly defining the process's starting point, various processing nodes (such as data extraction, conditional evaluation, and rule invocation), flow path, and endpoint. The process executor interprets and executes the defined process model. When the CRM platform generates an employee behavior event that meets the conditions, the process executor creates and runs a process instance, driving the data flow through the process according to the defined logical sequence. The task manager manages the various tasks involved in the process, including automated service tasks (such as calling external APIs for data enrichment), user tasks requiring manual intervention (such as approving specific actions), and rule tasks that invoke the rule engine module. A persistence mechanism saves the state of running process instances (such as the current node and variable values) to a data storage module, supporting long-running processes, asynchronous operations, and recovery from system failures.

[0088] More specifically, the rule engine module includes a rule base, a rule editor, an inference engine, and a working memory.

[0089] The rule base is used to centrally store all employee behavior evaluation rules defined by the dynamic configuration management module. These rules exist in a structured format (such as XML, JSON, and / or a specific rule language) to facilitate management and version control. The rule editor allows the business management end to create, view, modify, deploy, and manage rule sets. It can be part of the dynamic configuration management module or a standalone tool. The inference engine, as the brain of the rule engine, is responsible for matching and inferring based on the input facts (i.e., behavioral data from the process engine) and the rules in the rule base. Commonly used inference algorithms include the Rete algorithm and its variants (such as RetePlus), which improve matching efficiency by constructing optimized rule networks. During the execution of the inference engine, the input fact data is loaded into the working memory, and the rule matching operation is performed in the working memory.

[0090] Although this specification has shown and described several embodiments of the present invention, it is obvious to those skilled in the art that such embodiments are provided by way of example only. Those skilled in the art will conceive of many modifications, changes and substitutions without departing from the idea and spirit of the present invention.

Claims

1. A method for evaluating employee behavior based on a function engine, characterized in that: A CRM system for integrating a process engine and a rule engine, the method comprising: Define employee behavior rules in response to employee behavior configuration on the business management side; Invoke the process engine, encapsulate employee behavior data and context information into fact data, and transmit the fact data to the rule engine through the standardized interface of the CRM system; The rule engine is called to perform matching analysis on the fact data and the employee behavior rules to obtain an employee behavior score.

2. The employee behavior evaluation method based on the function engine according to claim 1 is characterized in that: The process engine is called to encapsulate the employee's behavior data and context information into fact data, specifically: In response to the generation of employee behavior data in the CRM system, capturing CRM events through an API interface or a message queue of the CRM system; According to the CRM event type, find and load the corresponding process engine; Importing the CRM event into the process engine to obtain a process instance; In the process instance, extracting employee behavior data and context information; The behavior data and the context information are encapsulated as fact data.

3. The employee behavior evaluation method based on the function engine according to claim 1 is characterized in that: The rule engine adopts a rule engine based on the Rete algorithm or the RetePlus algorithm.

4. The employee behavior evaluation method based on the function engine according to claim 3 is characterized in that: The rule engine is called to perform matching analysis on the fact data and the employee behavior rules to obtain an employee behavior score, specifically: In response to the generation of the fact data, the employee behavior rules are loaded into the Rete algorithm or the RetePlus algorithm and compiled into a directed acyclic graph matching network; The fact data is input into the directed acyclic graph matching network for data screening, data matching, conflicting data screening and scoring to obtain employee behavior scores.

5. The employee behavior evaluation method based on the function engine according to claim 4 is characterized in that: The directed acyclic graph matching network includes a root node, an object type screening node, an Alpha matching node, a Beta connection node and a terminal node.

6. The employee behavior evaluation method based on the function engine according to claim 5 is characterized in that: The fact data is input into the directed acyclic graph matching network for data screening, specifically: Using the root node as a network entry, distributing the fact data to a plurality of object type screening nodes; Distribute the object types that meet the preset screening rules in the object type screening node to the corresponding Alpha matching nodes.

7. The employee behavior evaluation method based on the function engine according to claim 4 is characterized in that: The fact data is input into the directed acyclic graph matching network for data matching, specifically: Using the Alpha matching node as the attribute matching node for a single fact data, the fact data that satisfies the employee behavior rule is stored in the Alpha memory; The Beta connection node is used as an attribute matching node that spans multiple fact data matches, and multiple fact data stored in the Alpha memory or output data of other Beta matching nodes are data connected to obtain a matching result.

8. The employee behavior evaluation method based on the function engine according to claim 7 is characterized in that: After obtaining the matching result, the method further includes: The matching result and / or any fact data of completing the evaluation of any employee behavior rule are transmitted to the corresponding terminal node.

9. The employee behavior evaluation method based on the function engine according to claim 8 is characterized in that: If multiple terminal nodes are activated at the same time, the fact data is input into the directed acyclic graph matching network to filter out conflicting data and score to obtain the employee behavior score, specifically: placing the matching results and / or fact data output by the multiple terminal nodes into an agenda queue for sorting, and filtering out duplicate matching results or duplicate fact data obtained based on the same employee behavior rule from the agenda queue; The matching results and / or fact data are scored item by item according to the order of the agenda queue to obtain the employee behavior score.

10. A CRM system, characterized in that: For implementing the employee behavior evaluation method based on the function engine according to any one of claims 1 to 9, the system comprises at least a dynamic configuration management module, a process engine module and a rule engine module; wherein, Dynamic configuration management module, used to respond to employee behavior configuration on the business management side and define employee behavior rules; A process engine module is used to encapsulate employee behavior data and context information into fact data and transmit the fact data to the rule engine module through the standardized interface of the CRM system; The rule engine module is used to match and analyze the fact data with the employee behavior rules to obtain an employee behavior score.

Images