Flexible employment matching method and system based on big data

Abstract

The invention relates to the technical field of human resource management, in particular to a flexible employment matching method and system based on big data, and the method comprises the steps: obtaining a post description text to establish behavior features, constructing a task association path to generate post intention elements, and building a flexible task distribution chart in combination with market demands; the job seeker information is integrated, the skill adaptation degree is evaluated, a skill matching list is generated, according to the skill matching list, resource allocation and job seeker behavior records are combined, a task combination with the optimal resource and time is screened out, and a flexible employment matching result is generated. According to the method, a behavior feature network is constructed by extracting post semantics, task intentions are aggregated based on logic paths, a task distribution system is constructed in combination with market demands, the skill matching degree is quantized through multi-dimensional evaluation, task paths are optimized through a resource and time balance model, and a closed loop of post demands, task extension and man-post matching is formed. And the matching timeliness, stability and collaboration are improved.

Description

Technical Field

[0001] This invention relates to the field of human resource management technology, and in particular to a flexible employment matching method and system based on big data. Background Technology

[0002] The field of human resource management technology encompasses various methods and means for the systematic management of labor acquisition, allocation, training, and utilization. Its core content is based on the supply and demand relationship in the labor market, and effectively organizes and allocates human resources through job analysis, recruitment processes, training systems, performance appraisals, and salary structures. The research and application of this technology not only involve the management of traditional employment positions, but also cover the management of personnel mobility, job diversity, and labor relations coordination under flexible employment models. It gradually forms a full-process human resource management system supported by informatization and data, thereby providing a standardized and scalable management framework for the employment needs of different organizations and individuals.

[0003] One such big data-based flexible employment matching method refers to a specific approach that analyzes and matches flexible workers using multi-source employment data and job information data. It primarily focuses on the supply and demand matching between the flexible employment group and temporary, short-term part-time opportunities. This involves collecting and comparing job seekers' basic information, work experience, skill characteristics, job requirements, task time requirements, and industry categories. The process typically includes collecting multi-dimensional data on job seekers and positions through data collection methods, organizing and summarizing information on different types of positions and personnel through data classification methods, and analyzing the correspondence between job seekers' skill tags and job task requirements through data comparison methods, thereby forming a basic methodology system for flexible employment matching.

[0004] Existing technologies in the flexible employment matching process mainly rely on static attribute comparison of job and personnel data, lacking in-depth analysis of job semantics and task behavior logic. As a result, job descriptions can only provide a superficial level of information. When there are logical breaks and semantic ambiguities between task objectives, the matching algorithm has difficulty identifying the potential connections and sequence between tasks. This causes the job intent to be unable to be accurately conveyed to the skill mapping layer, reducing the accuracy of job matching. Task allocation may result in resource waste and response delays, affecting the adaptation efficiency and execution continuity in flexible employment scenarios. Summary of the Invention

[0005] To address the technical problems existing in the prior art, embodiments of the present invention provide a flexible employment matching method and system based on big data. The technical solution is as follows: A flexible employment matching method based on big data includes the following steps: S1: Obtain action words, object words, and constraint words from the job description text, analyze the relationship between action words and object words, identify the performance forms in different task scenarios, combine the timeliness of paired constraint words with action goals, integrate behavioral patterns, and generate a set of job behavior features; S2: Based on the relationship between action words, object words and constraint words in the set of job behavior characteristics, group job task objectives, analyze the patterns of action words and object words, construct task objective association paths, identify extension directions and intersections, determine the logical order, and generate a set of job intent elements; S3: Call the task association information in the set of job intention elements, match task fragments in the job dataset and the set of employment demand tags, analyze the characteristics of the task fragments, determine the degree of fit between the task objectives and job requirements, establish the relationship between task intensity and region, and generate a flexible task distribution chart. S4: Based on the matching relationship between task objectives and job requirements in the flexible task distribution chart, extract skill information from the job seeker's skill library, analyze the suitability between skills and job requirements, assess the matching degree between skill proficiency and task requirements, and generate a skill matching list by combining task frequency and duration.

[0006] As a further aspect of the present invention, the set of job behavior characteristics includes behavior pattern characteristics, task target attributes, action and object association elements, constraint parameters, and scene performance characteristics; the set of job intent elements includes task target structure, logical association path, task connection relationship, and intent expression elements; the flexible task distribution chart includes task area division, task intensity level, job fit index, and demand tag elements; and the skill matching list includes skill type, proficiency level, job suitability index, and task correspondence.

[0007] As a further aspect of the present invention, the step of obtaining the job behavior feature set is as follows: S101: Obtain verb phrases and noun phrases from the job description text, detect the collocation of the action subject and the corresponding object in the sentence based on the syntactic structure, count the number of times each action word appears and the distribution ratio of the corresponding object word, calculate the difference in the frequency of action words and object words and record it in the data table, and generate action object association results; S102: Based on the action object association result, call the time modifiers and conditional restrictions in the job description text, compare the time interval and constraint conditions of the corresponding object words for each action word, determine the degree of matching between the time limit in the constraint and the action target, calculate the matching ratio and the extended offset interval, and obtain the action constraint matching result. S103: Based on the action constraint matching results, aggregate the paired content of all action words, object words and their constraint terms, identify the repetitive combination patterns of actions in different scenarios, calculate the number of aggregations and object crossover rates between the same action targets, and generate a set of job behavior features.

[0008] As a further aspect of the present invention, the step of obtaining the set of job intention elements is as follows: S201: Based on the relationship between action words, object words, and constraint words in the set of job behavior characteristics, analyze the frequency of occurrence of words, group task objectives, calculate the concentration of action words and object words in each task group, and generate task objective grouping results; S202: Based on the task target grouping results, analyze the repetition frequency of action words and the co-occurrence of object words in each task target, identify the co-occurrence patterns between task targets, extract high-frequency task association paths, and obtain the task target co-occurrence association results; S203: Based on the co-occurrence association results of the task objectives, extend the association paths between the task objectives, detect the logical connection order between the task objectives, sort the task objectives according to the intersection points of the paths and the extension direction, organize the association path information, and generate a set of job intention elements.

[0009] As a further aspect of the present invention, the step of obtaining the flexible task distribution chart is as follows: S301: Based on the task target association information in the set of job intent elements, extract relevant task fragments from the job description text, analyze the matching of action words and object words in each task fragment, and filter out task content that is highly related to job intent based on the relevance between task objectives and job requirements, and generate task fragment filtering results. S302: Based on the task segment screening results, obtain the connection between job task objectives and market demand, combine the key tags of job requirements, analyze the fit between task objectives and market demand, determine the degree of matching between task objectives and job requirements in each task segment, and obtain task demand matching results. S303: Based on the task requirement matching results, calculate the fit between the task objective and the job requirements, construct the correlation between task intensity and regional distribution based on the matching relationship between the task objective and the region, and generate a flexible task distribution chart.

[0010] As a further aspect of the present invention, the step of obtaining the skill matching list is as follows: S401: Based on the matching of task objectives and job requirements in the flexible task distribution chart, integrate the job seeker's past task records, certified skills and training information to establish a job seeker information database, extract skill information from the database, identify the relevance of each skill, and perform preliminary matching with job requirements to generate a set of job seeker skill information. S402: Based on the job seeker's skill information set, analyze the fit between each skill and the job requirements, compare the job seeker's skill proficiency with the job task requirements, and combine the frequency and duration of task execution to identify the fit between task objectives and job requirements, and obtain the task-skill matching degree. S403: Based on the task and skill matching degree, calculate the suitability between task objectives and job requirements, match the intensity of task objectives with task areas based on the suitability, organize and visualize task distribution, and generate a skill matching list.

[0011] As a further aspect of the present invention, the method further includes: S5: Based on the matching results of job seekers and job requirements in the skill matching list, extract the time consumption, resource demand and personnel response from the task path resource set and job seeker behavior record set, analyze the balance between resource demand and time allocation, select the optimal task combination based on the consumption-efficiency ratio, and generate flexible employment matching results. The flexible employment matching results include task combination schemes, resource allocation structures, time matching indicators, and response efficiency parameters.

[0012] As a further aspect of the present invention, the step of obtaining the flexible employment matching result is as follows: S501: Based on the matching status in the skill matching list, integrate the resource information involved in each task path, allocate resources according to the task objectives, calculate the correspondence between resource input and task quantity, form the resource distribution of the task path, and generate the task path resource set. S502: Based on the task path resource set, collect task completion time and efficiency data from the job seeker's past work records, combine their task participation frequency and response interval, calculate the execution time and completion rate of each task, organize the job seeker's work performance data, and establish a job seeker behavior record set. S503: Based on the time consumption, resource demand, and personnel response status of the job seeker behavior record set and the task path resource set, calculate the ratio of resource consumption to response efficiency, determine the balance between time and resource allocation, select task combinations with balanced resource allocation and high time matching degree, and generate flexible employment matching results.

[0013] A flexible employment matching system based on big data, the system comprising: The behavioral feature extraction module obtains action words, object words, and constraint words from the job description text, analyzes the relationship between action words and object words, identifies the performance forms in different task scenarios, and integrates behavioral patterns by combining the timeliness of paired constraint words with action goals to generate a set of job behavioral features. The task target association module, based on the relationship between action words, object words and constraint words in the set of job behavior characteristics, groups job task targets, analyzes the patterns of action words and object words, constructs task target association paths, identifies extension directions and intersections, determines the logical order, and generates a set of job intent elements. The demand matching and analysis module calls the task association information in the set of job intent elements, matches task fragments in the job dataset and the set of employment demand tags, analyzes the characteristics of the task fragments, judges the degree of fit between the task objectives and job requirements, establishes the relationship between task intensity and region, and generates a flexible task distribution chart. The skills matching assessment module extracts skills information from the job seeker's skills library based on the matching relationship between task objectives and job requirements in the flexible task distribution chart, analyzes the matching degree between skills and job requirements, assesses the matching degree between skill proficiency and task requirements, and generates a skills matching list by combining task frequency and duration. The matching result optimization module extracts time consumption, resource demand, and personnel response from the task path resource set and job requirement set based on the matching results of job seekers and job requirements in the skill matching list. It analyzes the balance between resource demand and time allocation, selects the optimal task combination based on the consumption-efficiency ratio, and generates flexible employment matching results.

[0014] The beneficial effects of the technical solutions provided in the embodiments of the present invention include at least the following: In this invention, by refining and mapping the semantics of job description texts, a computable semantic network is formed among actions, objects, and constraints based on task behavior features. Hierarchical connection and intent aggregation analysis of task objectives are achieved through logical path extension recognition. A task distribution system with regional association and intensity gradients is constructed through interactive comparison of task semantic features and dynamic market demands. A quantifiable skill matching system is formed by combining multi-dimensional evaluations of skill proficiency, task frequency, and execution time. Dynamic optimization of task paths is achieved through a resource and time balance allocation model, creating a closed loop between job requirement expression, task logical extension, and person-job matching, thereby improving the timeliness, stability, and execution synergy of the matching results. Attached Figure Description

[0015] Figure 1 This is a flowchart of the method of the present invention; Figure 2 This is a flowchart illustrating the process of obtaining the set of job behavior characteristics according to the present invention. Figure 3This is a flowchart illustrating the process of obtaining the set of job intent elements in this invention. Figure 4 This is a flowchart illustrating the process of obtaining the flexible task distribution chart of the present invention. Figure 5 This is a flowchart illustrating the process of obtaining the skill matching list in this invention. Figure 6 This is a flowchart illustrating the process of obtaining flexible employment matching results according to the present invention. Detailed Implementation

[0016] The technical solution of the present invention will now be described with reference to the accompanying drawings.

[0017] In embodiments of the present invention, words such as "exemplarily," "for example," etc., are used to indicate that something is an example, illustration, or description. Any embodiment or design described as "exemplary" in the present invention should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of the word "exemplary" is intended to present the concept in a concrete manner. Furthermore, in embodiments of the present invention, the meaning expressed by "and / or" can be both, or either one.

[0018] In the embodiments of this invention, the terms "image" and "picture" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning. Similarly, the terms "of," "corresponding (relevant)," and "corresponding" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning.

[0019] In this embodiment of the invention, sometimes a subscript such as W1 may be written in a non-subscript form such as W1. When the difference is not emphasized, the meaning they express is the same.

[0020] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0021] Please see Figure 1 This invention provides a technical solution: a flexible employment matching method based on big data, comprising the following steps: S1: Obtain the job description text, extract the action words, object words and constraint words in the description text, analyze the relationship between action words and object words, identify the performance in different task scenarios, combine the timeliness of constraint words with the action target to match, integrate the behavior pattern and establish job behavior characteristics, and generate a set of job behavior characteristics; S2: Based on the relationship between action words, object words, and constraint words in the job behavior feature set, group the various task objectives in the job, analyze the correspondence between action words and object words in each task objective, construct the association path between task objectives based on the recurrence of action words and the co-occurrence pattern between object words, identify the extension direction and intersection point of the association path, determine the logical connection order between task objectives, organize all association path information, and generate a set of job intent elements; S3: Call the task target association information in the job intent element set, extract task fragments related to job intent from the job description text, analyze the relationship between job task targets and market demand, obtain a set of employment demand tags, match them with job requirements in the employment demand tag set, identify key task content in task fragments, analyze the characteristics of each task fragment, determine the high degree of fit between task targets and job requirements, establish the matching relationship between task intensity and task area based on the degree of fit between task targets and job requirements, and generate a flexible task distribution chart.

[0022] S4: Based on the matching relationship between task objectives and job requirements in the flexible task distribution chart, integrate the job seeker's past task records, certification skills, and training information to generate a job seeker information database. Extract skill information from the job seeker information database, analyze the fit between each skill and job requirements, compare the job seeker's skill proficiency with the task requirements, and evaluate the job seeker's fit by combining the frequency and duration of task execution to generate a skill matching list.

[0023] S5: Based on the matching results between job seekers and job requirements in the skills matching list, integrate the resources required in each task path and allocate them according to the task objectives to obtain a task path resource set. Analyze the job seeker's past work performance, task completion time and efficiency to obtain a job seeker behavior record set. Identify time consumption, resource requirements and personnel response in the task path resource set and the job seeker behavior record set. Analyze the balance between various resources and time allocations. Combined with personnel response rate, based on the ratio of resource consumption to personnel response efficiency, select the task combination with the most balanced allocation of task resources and the highest time matching degree to generate flexible employment matching results.

[0024] The set of job behavior characteristics includes behavioral pattern characteristics, task goal attributes, action-object relationship elements, constraint parameters, and scenario performance characteristics. The set of job intent elements includes task goal structure, logical connection path, task connection relationship, and intent expression elements. The flexible task distribution chart includes task area division, task intensity level, job matching index, and demand tag elements. The skills matching list includes skill type, proficiency level, job suitability index, and task correspondence. The flexible employment matching results include task combination scheme, resource allocation structure, time matching index, and response efficiency parameters.

[0025] Please see Figure 2 The steps for obtaining the set of job behavior characteristics are as follows: S101: Obtain verb phrases and noun phrases from the job description text, detect the collocation of the action subject and the corresponding object in the sentence based on the syntactic structure, count the number of times each action word appears and the distribution ratio of the corresponding object word, calculate the difference in the frequency of action words and object words and record it in the data table, and generate action object association results; This method extracts verb phrases and noun phrases from job description texts and uses dependency parsing to analyze the job description for "Software Engineer" ("Design and implement high-performance backend services"). It identifies "design" and "implement" as verb phrases, and their corresponding noun phrases, "high-performance backend services," as object words. Based on syntactic structure, it detects the collocation of the action subject and its corresponding object, and statistically analyzes the frequency of each action word and its corresponding object word distribution ratio. For example, in ten "Software Engineer" job descriptions, "design" appears 80 times, "backend services" appears 60 times, and "database structure" appears 20 times, with the following distribution ratios: and The action word "implement" appeared 120 times, the object words "backend service" appeared 90 times, and "API interface" appeared 30 times, with the following distribution ratios: and The frequency difference between the action word and the object word is calculated and recorded in the data table. The frequency of the action word "design" is 80 times, and the frequency of the object word "backend service" is 60 times. The frequency difference is... The frequency of the action word "implement" is 120 times, and the frequency of the object word "backend service" is 90 times, with a frequency difference of 120 times. This data table is a matrix consisting of the frequency and difference of action words and their corresponding object words. The action words... With object word frequency difference The calculation uses the formula ,in, Action words Frequency of occurrence (number of times) Object words Frequency of occurrence (number of times) This represents the absolute value operation. In this embodiment, and The calculation is obtained by counting all job description texts. For example, in the job description of "software engineer" mentioned above, the frequency of the action word "design" is 80, and the frequency of the object word "backend service" is 60. Therefore... The frequency difference between the action word "design" and the object word "database structure" is 1. The frequency difference This reflects the degree of deviation between the overall emphasis of the action word mentioned in the text and the overall emphasis of the object word mentioned, and records it in the action-object association result table. For example, the association result of the action word "design" and its object "backend service" is (design, backend service, 80, 60, 20), and the association result of the action word "design" and its object "database structure" is (design, database structure, 80, 20, 60). This table contains all the pairings of action words and object words, their respective frequencies, and frequency differences, generating the action-object association result.

[0026] S102: Based on the action object association results, call the time modifiers and conditional restrictions in the job description text, compare the time interval and constraint conditions of the corresponding object words for each action word, determine the degree of matching between the time limit in the constraint and the action target, calculate the matching ratio and the extended offset interval, and obtain the action constraint matching results. Based on the action object association results, for the pairing of the action word "design" and its object word "backend service", time modifiers and conditional restrictions in the job description text are invoked, such as "to be completed within 3 months" and "to comply with". The "specification" compares the time frame and constraints of the corresponding object words for each action word. For the action word "design" and its object word "backend service," the constraint "complies with..." The "specification" is a conditional constraint with a timeframe of "to be completed within 3 months". The degree of match between the time constraint and the action objective is assessed. The time constraint "within 3 months" is matched with the action objective "design backend services". The time constraint is then used to calculate the degree of match. Standard time required to complete the action Matching ratio and extended offset interval Matching ratio The calculation formula is ,in, This represents the standard completion time of an action goal, such as the standard completion time for "designing backend services". The time limit is set at 4 months, with a requirement to "complete within 3 months". If it is 3 months, then The matching ratio The range of values ​​is The closer the value is to 1, the higher the matching degree, and the more extended the offset interval. The calculation formula is In this example, A negative value indicates that the time limit is shorter than the standard time. The matching ratio and extended offset interval are calculated to obtain the action constraint matching results, including the pairing of all action words and object words, as well as the corresponding time intervals, constraints, and matching ratios. and extended offset interval The results are recorded in the action constraint matching result table. Example data of the action constraint matching result is shown in Table 1. This table includes the pairing of action words and object words, time limit, standard completion time, matching ratio, and extended offset range. Table 1. Example of Action Constraint Matching Results As shown in Table 1, the matching ratio between the action word "design" and the object word "backend service" The offset range is 0.75. The value is -1 month, indicating that the time limit is one month shorter than the standard time, and the matching ratio between the action word "implementation" and the object word "API interface" is... A value of 0.8 indicates that the time limit of the task matches the standard time to a high degree. The matching results are then integrated to obtain the action constraint matching results.

[0027] S103: Based on the action constraint matching results, aggregate the paired content of all action words, object words and their constraint items, identify the repetitive combination patterns of actions in different scenarios, calculate the number of aggregations and object crossover rates between the same action targets, and generate a set of job behavior features. Based on the action constraint matching results, the paired content of all action words, object words, and their constraints is aggregated. Taking the pairing of the action word "design" and the object word "backend service" as an example, this pairing exists in different descriptive texts as "design backend service (within 3 months, following..."). The paired content is divided into two scenarios: "Specification" and "Designing backend services (within 2 months, high concurrency, high availability)". The system identifies repetitive combination patterns of actions in different scenarios, calculates the aggregation frequency and object crossover rate between identical action targets, and determines the aggregation frequency of the action target "Designing backend services". The crossover rate is 2 times. This represents the proportion of the number of elements in the intersection of the sets of object words that have the same action verb to the number of elements in the union of the sets of object words. There are two sets of object word pairings. and ,but In this example, the set of object words for the action goal "design" for Suppose another action goal is to "achieve" the set of object words. for ,but , , , ,therefore By analyzing action verbs Constraints in different scenarios and Compare and combine recurring constraints. As features, such as constraints The aggregation count is 1. Combining action words, object words, scenario constraint combinations, aggregation counts, and object crossover rates, a set of job behavior features is generated. Each feature in this set is (action word, object word, constraint combination, aggregation count, object crossover rate), for example (design, backend service). ).

[0028] Please see Figure 3 The steps to obtain the set of job intention elements are as follows: S201: Based on the relationship between action words, object words, and constraint words in the set of job behavior characteristics, analyze the frequency of word occurrence, group task objectives, calculate the concentration of action words and object words in each task group, and generate task objective grouping results; Based on the relationships between action words, object words, and constraint words in the set of job behavior characteristics, this study analyzes the frequency of word occurrence, groups task objectives, and pairs the action word "design" with its object words "backend service" and "database structure," as well as the action word "implementation" with its object words "backend service" and "database structure." The pairing of "interfaces" calculates the concentration of action words and object words within each task group. Concentration The calculation formula is: ; in and The number of action words and object words, respectively. Action words With object word The co-occurrence frequency is determined by setting a concentration threshold during the task objective grouping process. ,set up Next, group the task objectives, for example, by action words. With object word The co-occurrence frequencies are shown in Table 2. The concentration was calculated. Number of action words Number of object words The sum of co-occurrence frequencies ,but ,because This indicates that these words have a high degree of co-occurrence, and task objectives involving these words can be grouped into the same task group. If the calculation result is lower than Then the task objectives are divided into different task groups, generating task objective grouping results: ; The grouping is based on whether the co-occurrence concentration of action words and object words is higher than a threshold. ; Table 2. Co-occurrence frequency of action words and object words As shown in Table 2, the co-occurrence frequency of the action word "design" and the object word "backend service" is 60 times, and the co-occurrence frequency of the action word "implementation" and the object word "backend service" is also high. The co-occurrence frequency of the "interface" was 30 times; these values ​​were used to calculate the concentration within the task group. This concentration is higher than the threshold. Therefore, a task group was formed. Generate task target grouping results.

[0029] S202: Based on the task objective grouping results, analyze the repetition frequency of action words and the co-occurrence of object words in each task objective, identify the co-occurrence patterns between task objectives, extract high-frequency task association paths, and obtain the task objective co-occurrence association results; Grouping results based on task objectives Analyze the repetition frequency of action words and the co-occurrence of object words in each task objective, for example, in task groups. In the text, the action word "design" appears twice (designing backend services, designing database structure), and the object word "backend services" co-occurs in both the task objectives of "designing backend services" and "implementing backend services." This identifies the co-occurrence pattern between task objectives and calculates the results. and Co-occurrence frequency between Correlation coefficient with Pearson To construct association paths and correlation coefficients Used to measure the strength of the linear correlation between task objectives, its value range is: The closer the correlation is to 1, the stronger the positive correlation; the closer it is to -1, the stronger the negative correlation. Calculate the co-occurrence frequency of "designing backend services" and "implementing backend services": ; set up Next, let's assume an independent frequency for "designing backend services". The independent frequency of "implementing backend services" Next, calculate If the calculation result This indicates a strong positive correlation between the two. High-frequency task-related paths are extracted, and a co-occurrence frequency threshold for these high-frequency paths is set. Next, the determination of the paths connecting task objectives is based on and The conditions, such as "designing backend services" and "implementing backend services", must be met. and Therefore, the association path between the two is taken as the high-frequency task association path to obtain the task target co-occurrence association result, which includes all task target pairings that meet the conditions, their co-occurrence frequency and correlation coefficient.

[0030] S203: Based on the co-occurrence association results of task objectives, extend the association path between task objectives, detect the logical connection order between task objectives, sort the task objectives according to the intersection point of the path and the extension direction, organize the association path information, and generate a set of job intention elements. Based on the co-occurrence association results of task objectives, the association paths between task objectives are extended, and backend services are designed for high-frequency task association paths. "Implement backend services" involves detecting the logical connection order between task objectives, based on temporal logic analysis and causal relationship determination. For example, in all co-occurring texts, "design backend services" always appears before "implement backend services," and a temporal order weight is assigned. Causal weight Total weight ,in This represents the timing consistency ratio, in this example. , To determine the strength of the causal relationship, set ,but The task objectives are sorted based on the intersection points and extension directions of the paths. If a path "implements backend services" exists... "Testing service performance" and "Implementing backend services" are the points of intersection, with the extension direction being "Design". accomplish "Testing" determines the logical connection order between task objectives as (design backend service, implement backend service, test service performance), organizes the associated path information, and generates a set of job intent elements. This set includes data items such as task objective set, task group information, logical connection order between task objectives, co-occurrence frequency, and association strength. For example, (design backend service, implement backend service, 0.98) indicates that there is a high-strength logical connection order between the two.

[0031] Please see Figure 4 The steps to obtain a flexible task distribution chart are as follows: S301: Based on the task target association information in the job intent element set, extract relevant task fragments from the job description text, analyze the matching of action words and object words in each task fragment, and filter out task content that is highly related to the job intent based on the relevance between the task target and the job requirements, and generate task fragment filtering results. Based on the task objective association information in the job intent element set, relevant task fragments are extracted from the job description text. For the task objective "design backend services," the text fragment "responsible for..." is extracted. "Designing Backend Service Architecture under Standards" analyzes the matching of the action word "design" and the object word "backend service" in each task segment, based on the relevance of task objectives and job requirements. Select tasks that are highly relevant to the job objectives and set a relevance threshold. Correlation By comparing the overlap between keywords in the task segment and keywords in the job requirements. Keyword weight The weighted sum is calculated, for example, the keywords of the task fragment. Job Requirements Keywords Overlapping keywords There are 3 in total, and the total number of keywords required for the job postings is 4, with a high degree of overlap. Set keyword weight ,but Assuming semantic similarity ,but ,because The task segment is not relevant enough and will not be screened. However, the other segment, "Complete the design and delivery of high-performance services," is more relevant. If so, it will be filtered, and a highly relevant judgment interval will be set as follows: If the score is below 0.8, it is considered not highly relevant, and a task fragment screening result is generated. This result includes the screened task fragments and their relevance scores, for example (Complete the design and delivery of high-performance services, 0.85).

[0032] S302: Based on the task segment screening results, obtain the connection between job task objectives and market demand, combine the key tags of job requirements, analyze the fit between task objectives and market demand, judge the degree of matching between task objectives and job requirements in each task segment, and obtain the task requirement matching results. Based on the task fragment selection results, the relationship between job task objectives and market demands is obtained. For the selected task fragment "Complete the design and delivery of high-performance services," whose task objective is "design high-performance services," the degree of alignment between the task objective and market demands is analyzed by combining key tags of job requirements (such as "high concurrency"). Compatibility The coverage of market hot demands by the task objectives And the urgency weight of the demand The product is calculated as follows: Assuming the task objective of "designing high-performance services" covers the market hot demands of "microservice architecture" and "containerized deployment," The urgency weight of this demand If set to 0.9, then Determine the degree of match between the task objectives and job requirements in each task segment. Through calculation ,in ,but Set the range of highly compatible matching degree as follows: A score above 0.8 is considered a high match, yielding a task requirement matching result. This result includes the degree of matching between task segments, task objectives, and job requirements. For example, (Complete the design and delivery of high-performance services, 0.834) indicates that the task objectives are highly aligned with the job requirements.

[0033] S303: Based on the task requirement matching results, calculate the fit between task objectives and job requirements, construct the relationship between task intensity and regional distribution based on the matching relationship between task objectives and regions, and generate a flexible task distribution chart. Based on the task requirement matching results, the degree of fit between the task objectives and job requirements is calculated, and the degree of fit is determined accordingly. The task objective is "design high-performance services," and the task intensity is constructed based on the matching relationship between the task objective and the region. Regional distribution The relationship between them, task intensity The calculation formula is ,in Rate task complexity, such as "designing high-performance services". , Rate the duration of the task, for example ,but Regional distribution The matching relationship is calculated by determining the historical demand for the task objective in different regions. and regional salary levels weighted average Assuming the historical demand in region A is normalized to 0.9 and the salary level is normalized to 0.7, then... , increase task intensity and regional distribution By performing association and pairing, a flexible task distribution chart is generated. The chart is presented in the form of a heatmap, with the horizontal axis representing the region, the vertical axis representing the task objective, and the cell value representing the task intensity. For example, (Design high-performance service, Region A, 0.827) indicates task intensity. The mission objective is located in area A. The regional matching relationship is shown in the chart, which intuitively displays the distribution of task intensity for different task objectives in different regions.

[0034] Please see Figure 5 The steps to obtain the skill matching list are as follows: S401: Based on the matching of task objectives and job requirements in the flexible task distribution chart, integrate job seekers' past task records, certified skills and training information to establish a job seeker information database, extract skill information from the database, identify the relevance of each skill, and perform preliminary matching with job requirements to generate a set of job seeker skill information. Based on the matching between task objectives and job requirements in the flexible task distribution chart, the intensity of the task objective "design high-performance services" is determined. By integrating job seeker "Zhang San's" past task records, certified skills, and training information, a job seeker information database is established, and skill information is extracted from the database, such as the skills Zhang San possesses. Identify the relevance of each skill to job requirements (e.g., "high-concurrency service design"). By calculating the overlap between skill keywords and job requirement keywords. ,For example Relevance to "High-Concurrency Service Design" And conduct an initial matching with job requirements, and obtain a matching score. ,in ,but Preliminary matching score Assuming the skill weights required for the job The settings are based on the scarcity of skills in the market and the frequency of mention in job descriptions, with higher weights assigned to skills with high scarcity. Skill weights mentioned frequently In this example, It is considered a highly scarce skill, therefore it is set as follows: This generates a set of job seeker skills information, including each skill of the job seeker and its initial match score with the task objective, for example (Zhang San, ).

[0035] S402: Based on the job seeker's skill information set, analyze the fit between each skill and the job requirements, compare the job seeker's skill proficiency with the job task requirements, and combine the frequency and duration of task execution to identify the fit between task objectives and job requirements, and obtain the task-skill matching degree. Based on job seeker skill information sets, targeting skills " The initial matching score between "designing high-performance services" and the task objective was 0.63. Further analysis was conducted to assess the fit between each skill and the job requirements. By combining the job seeker's past project experience... and project duration Computational skills suitability Suppose Zhang San is in If you have 5 project experiences (normalized to 0.8) and 36 months of project duration (normalized to 0.9), then... Compare job seekers' skill proficiency The degree of matching with job requirements Proficiency Set to 4 (out of 5), the baseline proficiency level required for the task. Set to 3, matching degree: Combined with the frequency of task execution With duration For example, the frequency of the task "designing high-performance services" Next / Year, Duration Hours / year, identify the alignment between task objectives and job requirements. Compatibility The calculation formula is: Among them, task execution frequency The normalization process adopts ,in If the frequency is set to 20 times per year, which is the highest frequency, then... ,but: This yields the task-skill matching degree, which indicates the job seeker's skill suitability for a specific task objective.

[0036] S403: Based on the task and skill matching degree, calculate the fit between task objectives and job requirements, match the intensity of task objectives with task areas based on the fit degree, organize and visualize task distribution, and generate a skill matching list. Based on task and skill matching Calculate the fit between task objectives and job requirements. Adaptability Compatibility Maintain consistency, that is The intensity of the task objective is determined based on the suitability. With the mission area Matching based on factors such as task intensity. Regional distribution of region A Matching metrics ,but Organize and visualize the task distribution to generate a skills matching list, which includes job seekers, task objectives, skills, and suitability. and matching metrics Information such as (Zhang San, designing high-performance services, etc.) ), this matching metric The range of values ​​is A higher value indicates a greater overall match between the job seeker and the task objective in terms of skills, intensity, and location.

[0037] Please see Figure 6 The steps to obtain flexible employment matching results are as follows: S501: Based on the matching status in the skill matching list, integrate the resource information involved in each task path, allocate resources according to the task objectives, calculate the correspondence between resource input and task quantity, form the resource distribution of the task path, and generate the task path resource set. Based on the matching results between job seeker "Zhang San" in the skills matching list and the task "Design high-performance services" Integrate resource information involved in each task path, and determine the computing resources required for the task "designing high-performance services". Utility software and collaborators ,set up , , Resources are allocated based on task objectives, and the correspondence between resource input and the number of tasks is calculated. For example, a task called "Designing High-Performance Services" requires the above-mentioned resource input. We use a weighted summation of resource consumption, and set weights. The resource quantities are normalized. Assume the normalized resource quantities are as follows: ,but This process creates a resource distribution along the task path, generating a task path resource set. This set includes the task objective, the required resource types, the resource quantities, and the normalized resource input. For example (designing high-performance services, ).

[0038] S502: Based on the task path resource set, collect task completion time and efficiency data from job seekers' past work records, combine their task participation frequency and response interval, calculate the execution time and completion rate of each task, organize them into job seekers' work performance data, and establish a job seeker behavior record set. Based on the task path resource set, and targeting job seeker "Zhang San" and the task objective "design high-performance service", task completion times from the job seeker's past work records are collected. With efficiency data For example, Zhang San's average completion time on similar tasks Hourly average efficiency Combined with their task participation frequency and response interval ,For example Times / year, average response interval Calculate the execution time of each task within hours. With completion rate Execution time The calculation formula is: ; Completion rate The job performance data of job seekers was compiled and a job seeker behavior record set was established. Part of the data in the job seeker behavior record set is shown in Table 3. This set includes job seekers, task objectives, and actual completion time. ,efficiency Execution time and completion rate Data items, such as (Zhang San, designing high-performance services, etc.) ); Table 3: Partial Record of Job Seeker "Zhang San's" Behavioral Characteristics As shown in Table 3, the execution time of job seeker "Zhang San" on the task "Design a high-performance service" is... for Hours, completion rate for Task / hour: Execution time of the task "Implement API Interface" for Hours, completion rate for Tasks per hour are used for subsequent resource allocation balance calculations.

[0039] S503: Based on the time consumption, resource demand, and personnel response of the job seeker behavior record set and the task path resource set, calculate the ratio of resource consumption to response efficiency, judge the balance between time and resource allocation, screen out task combinations with balanced resource allocation and high time matching degree, and generate flexible employment matching results. Based on the time occupancy of job seeker behavior records and task path resource sets Resource requirements and personnel response status Time commitment This refers to the execution time of the task. resource demand That is, normalized resource input. Personnel response status By response interval reciprocal The number of times per hour represents the ratio of computational resource consumption to response efficiency. ,ratio The calculation formula is ,in Resource consumption per unit of time For the sake of personnel response efficiency, in this example, The calculation process is as follows To determine the balance between time and resource allocation, the balance index is... Select task combinations with the most balanced resource allocation and highest time matching, and set the resource balance level. The benchmark value Time matching degree The benchmark value Time matching degree pass Calculate, assuming the task objective completion time ,but The calculation process is as follows The filtering criteria are: and In this example, and The task combination (Zhang San, design high-performance services) meets the screening criteria, generating flexible employment matching results. These results include successfully matched job seekers, task objectives, and job balance. Time matching degree For example (Zhang San, designing high-performance services, The results indicate that the task combination has a high degree of balance and matching in terms of resource allocation and time matching. This result shows that the ratio of resource consumption to personnel response efficiency of the task combination is extremely low, and its resource balance is close to 1.0, reaching a high degree of balance.

[0040] A flexible employment matching system based on big data, the system includes: The behavioral feature extraction module obtains action words, object words, and constraint words from the job description text, analyzes the relationship between action words and object words, identifies the performance forms in different task scenarios, and integrates behavioral patterns by combining the timeliness of paired constraint words with action goals to generate a set of job behavioral features. The task objective association module, based on the relationship between action words, object words, and constraint words in the set of job behavior characteristics, groups job task objectives, analyzes the patterns of action words and object words, constructs task objective association paths, identifies extension directions and intersections, determines the logical order, and generates a set of job intent elements. The demand matching and analysis module calls the task association information in the job intent element set, matches task fragments in the job dataset and the employment demand tag set, analyzes the characteristics of the task fragments, judges the degree of fit between the task objectives and job requirements, establishes the relationship between task intensity and region, and generates a flexible task distribution chart. The skills matching assessment module extracts skills information from job seekers' skills base based on the matching relationship between task objectives and job requirements in the flexible task distribution chart, analyzes the matching degree between skills and job requirements, assesses the matching degree between skill proficiency and task requirements, and generates a skills matching list by combining task frequency and duration. The matching result optimization module extracts time consumption, resource demand, and personnel response from the task path resource set and job requirement set based on the matching results of job seekers and job requirements in the skill matching list. It analyzes the balance between resource demand and time allocation, selects the optimal task combination based on the consumption-efficiency ratio, and generates flexible employment matching results.

[0041] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A flexible employment matching method based on big data, characterized in that, The method comprises the following steps: S1: obtaining action words, object words and constraint words in the post description text, analyzing the relationship between the action words and the object words, identifying the manifestation in different task scenarios, integrating the behavior mode, combining the timeliness of the paired constraint words and the action target, generating a post behavior feature set; S2: grouping post task targets based on the relationship between the action words, object words and constraint words in the post behavior feature set, analyzing the mode of the action words and the object words, constructing a task target association path, identifying the extension direction and intersection, determining the logical sequence, and generating a post intent element set; S3: calling the task association information in the post intent element set, matching task fragments in the post data set and the labor demand label set, analyzing the characteristics of the task fragments, judging the fit degree of the task target and the post requirements, establishing the task intensity and regional relationship, and generating a flexible task distribution chart; S4: according to the matching relationship between the task target and the post requirements in the flexible task distribution chart, extracting skill information from the job seeker skill library, analyzing the adaptability of the skill and the post demand, evaluating the matching degree of the skill proficiency and the task requirement, combining the task frequency and the time length, and generating a skill matching list.

2. The flexible employment matching method based on big data according to claim 1, characterized in that: The post behavior feature set includes behavior mode features, task target attributes, action and object association elements, constraint condition parameters and scene performance features, the post intent element set includes task target structures, logical association paths, task connection relationships and intent expression elements, the flexible task distribution chart includes task area division, task intensity level, post fit index and demand label elements, and the skill matching list includes skill type, proficiency level, post adaptation index and task corresponding relationship. 3.The flexible employment matching method based on big data according to claim 1, characterized in that: The acquisition step of the post behavior feature set is: S101: obtaining verb phrases and noun phrases in the post description text, detecting the collocation of action subjects and corresponding objects in the sentence according to the syntactic structure, counting the number of occurrences of each action word and the distribution ratio of the corresponding object word, calculating the frequency difference between the action word and the object word and recording it in the data table, and generating an action-object association result; S102: based on the action-object association result, calling the time modifiers and conditional limiting words in the post description text, comparing the time effectiveness interval and the constraint condition of the object word corresponding to each action word, judging the matching degree between the time limit item in the constraint word and the action target, calculating the matching proportion and the extension offset interval, and obtaining the action constraint matching result; S103: according to the action constraint matching result, aggregating the pairing content of all action words, object words and constraint items, identifying the repeated combination mode of the action in different scenarios, calculating the aggregation times and object intersection rate between the same action targets, and generating a post behavior feature set.

4. The big data based flexible employment matching method according to claim 1, characterized in that: The acquisition step of the post intent element set is: S201: based on the relationship between the action words, object words and constraint words in the post behavior feature set, analyzing the frequency of the words, grouping the task targets, calculating the concentration degree of the action words and the object words in each task group, and generating a task target grouping result; S202: According to the task target grouping result, the repetition number of action words and the co-occurrence of object words in each task target are analyzed, the co-occurrence mode between task targets is identified, the high-frequency task association path is extracted, and the task target co-occurrence association result is obtained; S203: Based on the task target co-occurrence association result, the association path between task targets is extended, the logical connection sequence between task targets is detected, the task targets are sorted according to the path intersection point and the extension direction, the association path information is arranged, and the post intent element set is generated.

5. The big data based flexible employment matching method according to claim 1, characterized in that: The acquisition step of the flexible task distribution chart is: S301: According to the task target association information in the post intent element set, the related task fragments in the post description text are extracted, the matching of action words and object words in each task fragment is analyzed, the task content highly related to the post intent is selected according to the relevance of task targets and post requirements, and the task fragment selection result is generated; S302: Based on the task fragment selection result, the relationship between post task targets and market demand is obtained, the matching degree between task targets and market demand is analyzed by combining the key labels of post requirements, the matching degree of task targets in each task fragment and post requirements is judged, and the task demand matching result is obtained; S303: According to the task demand matching result, the matching degree of task targets and post requirements is calculated, the association between task intensity and regional distribution is constructed according to the matching relationship between task targets and regions, and the flexible task distribution chart is generated.

6. The big data based flexible employment matching method according to claim 1, characterized in that: The acquisition step of the skill matching list is: S401: According to the matching of task targets and post requirements in the flexible task distribution chart, the past task records, certified skills and training information of job seekers are integrated, a job seeker information library is established, the skill information in the library is extracted, the relevance of each skill is identified, and a preliminary matching between the skill and the post demand is performed, and a job seeker skill information set is generated; S402: Based on the job seeker skill information set, the adaptation between each skill and post demand is analyzed, the matching degree of skill proficiency of job seekers and post task requirements is compared, the matching degree of task targets and post requirements is identified by combining the frequency and length of task execution, and the task and skill matching degree is obtained; S403: According to the task and skill matching degree, the adaptation degree of task targets and post requirements is calculated, the intensity of task targets and the matching of task regions are matched according to the adaptation degree, the task distribution is arranged and visualized, and the skill matching list is generated. 7.The flexible employment matching method based on big data according to claim 1, characterized in that: The method further comprises: S5: According to the matching result of job seekers and post requirements in the skill matching list, the time occupation, resource demand and personnel response in the task path resource set and the job seeker behavior record set are extracted, the balance of resource demand and time allocation is analyzed, the optimal task combination is selected according to the consumption and efficiency ratio, and the flexible employment matching result is generated; The flexible employment matching result includes task combination scheme, resource allocation structure, time matching index and response efficiency parameter. 8.The flexible employment matching method based on big data according to claim 7, characterized in that: The acquisition step of the flexible employment matching result is: S501: According to the matching in the skill matching list, integrate the resource information involved in each task path, allocate resources according to the task target, calculate the corresponding relationship between resource input and task quantity, form the resource distribution of the task path, and generate the task path resource set; S502: Based on the task path resource set, collect the task completion time and efficiency data in the past work records of job seekers, combine their task participation frequency and response interval, calculate the execution time and completion rate of each task, organize the job seeker's work performance data, and establish the job seeker behavior record set; S503: According to the time occupation, resource demand and personnel response in the job seeker behavior record set and the task path resource set, calculate the ratio of resource consumption and response efficiency, judge the balance degree of time and resource allocation, select the task combination with balanced resource allocation and high time matching degree, and generate the flexible employment matching result.

9. A flexible employment matching system based on big data, characterized in that, The system is used for the flexible employment matching method based on big data in any one of claims 1-8, and the system comprises: a behavior feature extraction module that obtains action words, object words and constraint words in a job description text, analyzes the relationship between action words and object words, identifies the manifestation in different task scenarios, integrates behavior patterns in combination with the timeliness of matching constraint words and action targets, and generates a job behavior feature set; a task target association module that groups job task targets based on the relationship between action words, object words and constraint words in the job behavior feature set, analyzes the mode of action words and object words, constructs a task target association path, identifies the extension direction and intersection, determines the logical sequence, and generates a job intent element set; a demand matching analysis module that calls the task association information in the job intent element set, matches task fragments in the job data set and the labor demand tag set, analyzes the characteristics of the task fragments, judges the fit degree of the task target and the job requirements, establishes the task intensity and regional relationship, and generates a flexible task distribution chart; a skill adaptation evaluation module that extracts skill information from a job seeker skill library according to the matching relationship between task targets and job requirements in the flexible task distribution chart, analyzes the adaptation degree of skills and job requirements, evaluates the matching degree of skill proficiency and task requirements, and generates a skill matching list in combination with task frequency and duration; a matching result optimization module that extracts time occupation, resource demand and personnel response from the task path resource set and the job seeker behavior record set according to the matching result of job seekers and job requirements in the skill matching list, analyzes the balance of resource demand and time allocation, selects the optimal task combination according to the consumption and efficiency ratio, and generates a flexible employment matching result.

Images