Intelligent matching and recommendation system for zero workers based on multi-dimensional dynamic label system

Abstract

The application discloses a zero worker post intelligent matching and recommendation system based on a multi-dimensional dynamic label system, and relates to the technical field of zero worker post intelligent matching.The application first collects zero worker job-seeking and enterprise recruitment multi-source data through a label construction module, and constructs six-dimensional one-to-one correspondence zero worker dynamic labels and post dynamic labels;then a post matching module establishes a dimensional label group according to the dimensional correspondence relationship, calculates the similarity of each dimension by using the cosine similarity after weighting, and obtains the basic matching degree by weighting; subsequently, the comprehensive matching score is calculated by combining the supply and demand and cost factors, and the optimal combination is optimized according to the comprehensive matching score to complete the matching; the application solves the problems of traditional static, single-dimensional label description incompleteness and update lag from the label source, realizes the accurate description of the explicit and implicit attributes of zero workers and posts, greatly reduces the post mismatch rate, and significantly improves the matching accuracy and efficiency.

Description

Technical Field

[0001] This invention belongs to the field of intelligent matching technology for zero-worker positions, specifically an intelligent matching and recommendation system for zero-worker positions based on a multi-dimensional dynamic tagging system. Background Technology

[0002] With the rapid development of the gig economy and the continuous growth in demand for flexible employment, gig worker job matching has become a core component of the flexible employment service sector. Most existing gig worker job matching technologies employ static tagging systems for describing job characteristics and calculating similarity. These typically only set tags around a single dimension or a few fixed dimensions such as age, education, and skill type. The tag content is fixed and outdated, failing to reflect dynamic information such as gig worker skill development, changes in performance, and adjustments in job requirements in real time. Furthermore, existing matching methods often directly use simple rule matching or basic similarity calculations, leading to significant deviations in similarity calculation results due to issues such as mismatched tag dimensions and incomplete feature characterization.

[0003] This invention provides an intelligent matching and recommendation system for zero-worker positions based on a multi-dimensional dynamic tagging system to solve the above-mentioned technical problems. Summary of the Invention

[0004] The present invention aims to solve at least one of the technical problems existing in the prior art; to this end, the present invention proposes a zero-worker job intelligent matching and recommendation system based on a multi-dimensional dynamic tagging system.

[0005] To achieve the above objectives, the first aspect of the present invention provides a zero-worker job intelligent matching and recommendation system based on a multi-dimensional dynamic tagging system, including a job matching module and a tagging construction module; Tag building module: used to build multi-dimensional dynamic tags; among which, multi-dimensional dynamic tags include gig dynamic tags and job dynamic tags, as well as the dimensional correspondence between gig dynamic tags and job dynamic tags; The job matching module is used to match job similarity based on multi-dimensional dynamic tags to obtain job matching groups; it calculates the comprehensive matching score of each matching combination in the job matching group, and selects the optimal matching combination based on the comprehensive matching score to achieve matching of casual workers and job positions.

[0006] In one possible implementation, the dynamic tags for gig workers include basic attribute tags, skill attribute tags, behavioral fulfillment tags, scenario adaptation tags, salary preference tags, and real-time status tags; the dynamic tags for job positions include basic requirement tags, skill requirement tags, fulfillment requirement tags, scenario attribute tags, salary and benefits tags, and real-time requirement tags.

[0007] One possible implementation involves matching job similarity based on multi-dimensional dynamic tags, including: Dimensional label groups are constructed based on the dimensional correspondence of multi-dimensional dynamic labels, and weight coefficients are set for each dimensional label group. Calculate the tag similarity of each dimension of the tag group; perform a weighted sum based on the weight coefficient and the tag similarity to obtain the basic matching degree; where the tag similarity is calculated using cosine similarity. The system retrieves a preset matching threshold and generates job-person matching groups based on the base matching score and the matching threshold. The default value for the matching threshold is 0.6.

[0008] In one possible implementation, the comprehensive matching score for each matching combination in the job-person matching group is calculated, including: The supply and demand imbalance index is used to set weights for supply and demand adjustment and labor cost; the supply and demand imbalance index is used to characterize the degree of supply and demand imbalance in the gig economy. The comprehensive matching score of each matching combination in the job matching group is calculated based on the supply and demand adjustment weight and the labor cost weight; the comprehensive matching score is used to quantify the adaptation priority of the matching combination.

[0009] In one possible implementation, supply and demand adjustment weights and labor cost weights are set based on the supply and demand imbalance index, including: The supply-demand imbalance index is input into a pre-constructed supply-demand adjustment function to obtain the supply-demand adjustment weight; where the higher the supply-demand imbalance index, the lower the corresponding supply-demand adjustment weight. The labor cost weight is obtained by solving the problem using the sum of the supply and demand adjustment weight and the labor cost weight, which is 1.

[0010] In one possible implementation, the comprehensive matching score for each matching combination in the job-person matching group is calculated based on supply-demand adjustment weights and labor cost weights, including: Calculate the labor cost of the matched combination; where labor cost includes food cost and commuting cost; Based on the weights of supply and demand adjustment and labor cost, the basic matching degree of the matching combination and the labor cost are weighted and summed to obtain the comprehensive matching score.

[0011] In one possible implementation, a supply-demand imbalance index is calculated based on job supply and demand data, including: Retrieve job and labor supply and demand data for the target area; the target area is the region that needs matching and recommendation of gig workers, and the job and labor supply and demand data includes the number of gig workers and positions in the target area; The supply-demand imbalance index of several target sub-regions in the target region is calculated based on the supply and demand data of personnel and positions; the supply-demand imbalance index is the ratio of the number of casual workers to the number of job positions in the target sub-region.

[0012] One possible implementation involves optimizing the job-person matching group based on the supply-demand imbalance index, including: The matching degree adjustment ratio of the target sub-region is based on the supply and demand imbalance index; wherein the preset range of the matching degree adjustment ratio is [-0.2, 0.2]; The basic matching degree of the matching combination associated with the job in the target sub-region is adjusted based on the matching degree adjustment ratio to optimize the person-job matching group.

[0013] In one possible implementation, the matching ratio of the target sub-region is adjusted based on the supply-demand imbalance index, including: Retrieve the pre-constructed matching degree adjustment function; where the matching degree adjustment function is a linear function; By substituting the supply-demand imbalance index of the target sub-region into the matching degree adjustment function, the matching degree adjustment ratio of the target sub-region is obtained.

[0014] In one possible implementation, optimizing the job-person matching group based on labor costs includes: Calculate the labor cost of each matching combination in the job matching group; Retrieve the preset cost cap for gig workers; compare the labor cost with the cost cap to delete unsuitable matching combinations, thereby optimizing the job-person matching group.

[0015] Compared with the prior art, the beneficial effects of the present invention are: 1. This invention first collects multi-source data on gig job seekers and corporate recruitment through a tag construction module, constructing dynamic gig job tags and job dynamic tags with six dimensions corresponding one-to-one. Then, a person-job matching module assembles dimensional tag groups based on dimensional correspondences, assigns weights, and calculates the similarity of each dimension using cosine similarity, weighting the results to obtain a basic matching degree. Subsequently, a comprehensive matching score is calculated by combining supply and demand and cost factors, and the optimal combination is selected based on the comprehensive matching score to complete the matching. This invention solves the problems of incomplete characterization and delayed updates of traditional static, single-dimensional tags from the source of tags, achieving accurate characterization of the explicit and implicit attributes of gig jobs and jobs, significantly reducing the mismatch rate between people and jobs, and significantly improving matching accuracy and efficiency. Dynamic tags adapt to the flexible and high-frequency characteristics of the gig job market, and can respond in real time to changes such as skill upgrades and demand changes, ensuring that the matching results are always in line with reality. The comprehensive score quantifies the matching priority, solving the problem of multiple combination sorting, ensuring matching quality, simplifying the matching decision-making process, and providing support for gig workers to quickly start work and companies to efficiently employ workers.

[0016] 2. After generating job-person matching groups, this invention first calculates a supply-demand imbalance index based on job-person supply and demand data. A matching degree adjustment function is then used to adjust the basic matching degree within a preset range to balance regional labor supply and demand. Next, the labor cost of each matching combination is calculated and compared with the preset cost ceiling for gig workers. Combinations exceeding cost expectations are eliminated, completing a dual optimization of the matching groups. After optimization, the supply-demand adjustment weight and labor cost weight are determined through the supply-demand imbalance index, and a weighted comprehensive matching score is calculated to ultimately determine the optimal match. This invention incorporates market supply and demand fluctuations and the actual labor cost of gig workers into the matching optimization process, using data-driven methods to address the shortcomings of traditional matching methods that neglect regional supply-demand imbalances and gig worker costs. Addressing the pain point of excessively high labor costs, the system employs a dual optimization approach, simultaneously improving matching quality and reducing computational burden. By adjusting the matching degree through a supply-demand imbalance index, it effectively alleviates the resource mismatch problem of oversupply and undersupply of gig workers in the region, achieving a balance between labor supply and demand across the entire region. Labor cost screening aligns with the actual employment needs of gig workers, reducing their burden and increasing their willingness to accept orders and their employment experience. Pre-optimization eliminates invalid combinations, reducing subsequent computation and improving system efficiency. Ultimately, it achieves a multi-objective synergy of matching accuracy, regional supply-demand balance, and controllable gig worker costs, balancing enterprise labor efficiency with gig worker employment experience, and promoting the efficient and stable operation of the gig worker market. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a schematic diagram illustrating the working principle of the zero-worker intelligent matching and recommendation system in an embodiment of the present invention. Figure 2 This is a schematic diagram of the workflow for generating person-job matching groups based on multi-dimensional dynamic tags in an embodiment of the present invention; Figure 3 This is a schematic diagram of the matching and combination optimization process in an embodiment of the present invention. Detailed Implementation

[0019] The technical solution of the present invention will be clearly and completely described below with reference to the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0020] Please see Figure 1The first aspect of this invention provides a smart matching and recommendation system for gig workers and jobs based on a multi-dimensional dynamic tagging system, including a person-job matching module and a tag building module; the tag building module is used to build multi-dimensional dynamic tags; wherein, the multi-dimensional dynamic tags include gig worker dynamic tags and job dynamic tags, as well as the dimensional correspondence between gig worker dynamic tags and job dynamic tags; the person-job matching module is used to perform person-job similarity matching based on the multi-dimensional dynamic tags to obtain person-job matching groups; calculate the comprehensive matching score of each matching combination in the person-job matching group, and realize the matching of gig workers and jobs based on the comprehensive matching score; wherein, the matching combination includes a gig worker and a job.

[0021] The job matching module and the tag building module interact with each other, and they also interact with the system's data storage module to obtain other data required for job matching and recommendation.

[0022] Existing gig worker job matching solutions often employ single-dimensional or limited-dimensional static tags, failing to reflect real-time changes in gig worker skills, job demand, and market fluctuations. Furthermore, mismatched tag dimensions hinder effective similarity calculations, leading to mismatches and low matching efficiency. This invention constructs multi-dimensional dynamic tags that comprehensively cover six corresponding dimensions of gig workers and job positions: basic skills, behaviors, and scenarios. These tags accurately depict both explicit and implicit attributes, and a multi-trigger dynamic update mechanism ensures tag timeliness, resolving the disconnect between static tags and actual conditions. This provides reliable data support for subsequent basic matching calculations, supply and demand optimization, and commuting cost optimization, laying the foundation for accurate job matching. Ultimately, this helps improve matching accuracy, balance regional supply and demand, and reduce gig worker commuting costs and enterprise labor costs.

[0023] Tag building module: Used to collect multi-source data to construct multi-dimensional dynamic tags; multi-dimensional dynamic tags include gig worker dynamic tags and job posting dynamic tags. Multi-source data includes job application data uploaded by gig workers and recruitment data uploaded by companies.

[0024] The dynamic tags for gig workers include basic attribute tags, skill attribute tags, behavioral fulfillment tags, scenario adaptation tags, salary preference tags, and real-time status tags; The basic attribute tags are static tags, and the tag items include age, education, etc. They are mainly used to describe the basic qualifications for temporary workers, corresponding to the basic requirement tags in the dynamic tags of the job position. The skill attribute tags are dynamic tags. The tag items include skill type (such as housekeeping, logistics, manufacturing, etc.), skill proficiency, qualification certificates, etc., which are used to describe the skill level of casual workers and correspond to the skill requirement tags in the dynamic tags of the job positions. The performance tags are dynamic tags, and the tag items include performance rate, on-time arrival rate, evaluation score, order rejection rate, order acceptance frequency, etc., which are used to characterize the performance capabilities and service quality of gig workers, and correspond to the performance requirements tags of the job dynamic tags; The scenario adaptation tag is a dynamic tag, and the tag items include work intensity adaptability, environmental adaptability, industry scenario adaptation experience, etc., which are used to describe the adaptability of gig workers to job scenarios, corresponding to the scenario attribute tags in the job dynamic tags. The salary preference tag is a dynamic tag, and the tag items include salary expectations, payment methods, benefits, etc., which are used to characterize the demand for gig employment and correspond to the salary and benefits tags in the dynamic tags of the job. The real-time status label is a dynamic label, and the label items include regional preferences, current location, and available time for accepting orders. It is used to depict the real-time matchability status of casual workers and corresponds to the real-time demand label in the dynamic label of the job position.

[0025] Job dynamic tags include basic requirements tags, skill requirements tags, performance requirements tags, scenario attribute tags, salary and benefits tags, and real-time requirements tags; The basic requirement tags are static tags, and the tag items include age, education, etc., which are used to describe the basic requirements of the job. The skill requirement tags are dynamic tags, and the tag items include skill requirement type, skill proficiency requirements, qualification requirements, etc., which are used to describe the skill requirements of the job. The performance requirement labels are dynamic labels, and the label items include performance rate requirements, on-time arrival requirements, service quality requirements, etc., which are used to describe the job requirements for the performance capabilities of gig workers. The scenario attribute tags are dynamic tags, and the tag items include work intensity, work environment, industry scenario, etc., which are used to describe the characteristics of the job scenario. The salary and benefits tag is a dynamic tag, and the tag items include salary standards, payment methods, benefits, etc., which are used to describe the salary advantages of the position. Real-time demand tags are dynamic tags, and the tag items include urgency, work location, work hours, and remaining labor shortage, which are used to depict the real-time labor demand for the position.

[0026] It is worth noting that a real-time update mechanism has been established for multi-dimensional dynamic tags to ensure that these tags remain consistent with the actual status of gig workers and job positions. The following is an example illustrating the real-time update mechanism: 1) Behavior trigger: When a gig worker completes order acceptance, fulfillment, and evaluation, or when the company updates job requirements or job evaluation, the tag items in the multi-dimensional dynamic tags will be automatically updated.

[0027] 2) Event Trigger: When gig workers acquire new qualification certificates or upgrade their skills, or when companies experience changes in job requirements or urgent staffing, the tag item update will be triggered manually or automatically.

[0028] 3) External data triggers: Based on changes in external traffic conditions and market demand data, update the real-time status tags of gig workers and the real-time demand tags of job positions.

[0029] In a preferred embodiment, please refer to Figure 2 Based on multi-dimensional dynamic tags, the similarity between people and positions is matched to obtain the matching groups.

[0030] Multi-dimensional dynamic tags accurately construct multi-dimensional tag groups for gig workers and job positions, comprehensively and in real-time depicting the attribute characteristics of gig workers and job positions, providing a standardized and quantifiable core basis for basic matching degree calculation. Next, the person-job matching module reasonably assigns weight coefficients to each dimension tag group and tag item group, calculates the tag similarity of each dimension tag group by combining cosine similarity, and then obtains an accurate basic matching degree through weighted summation. Subsequently, person-job matching groups are generated by filtering through preset matching degree thresholds. This ensures the accuracy and rationality of matching, and highlights the importance of core dimensions through weight allocation, avoiding invalid matching. This lays a solid foundation for subsequent optimization based on the supply and demand of people and jobs and commuting costs, further improving the efficiency and quality of person-job matching, while adapting to the flexible and high-frequency characteristics of the gig market.

[0031] The workflow for constructing dimension label groups based on dimension correspondence and then performing person-job similarity matching to obtain person-job matching groups is as follows: J01: Construct dimension label groups based on the dimension correspondence of multi-dimensional dynamic labels, and set weight coefficients for each dimension label group; The weighting coefficients can be based on the analytic hierarchy process (AHP), combined with the priority of dynamic tags for gig workers or job positions, to assign weights to each dimension tag group. Dimension tag groups are constructed based on the dimensional correspondence between tags, and include six groups: basic attribute tags and basic requirement tags, skill attribute tags and skill requirement tags, behavioral fulfillment tags and fulfillment requirement tags, scenario adaptation tags and scenario attribute tags, salary preference tags and salary and benefits tags, and real-time status tags and real-time requirement tags.

[0032] For example, if the skill dimension has the highest priority in the gig work dynamic tag or job dynamic tag, then the weight coefficient of the corresponding dimension tag group of the skill attribute tag and skill requirement tag is the highest, such as set to 0.25, and the weight coefficient of the other dimension tag groups is set to 0.15.

[0033] J02: Calculate the tag similarity of tag groups in each dimension; The label similarity of each dimension label group is calculated using cosine similarity, which involves substituting the quantified values ​​of each label item in the dimension label group into the cosine similarity calculation formula and then summing them by weight to obtain the label similarity.

[0034] A tag group is a group constructed from the corresponding tag items in a dimension tag group. For example, the skill type in the skill attribute tag and the skill requirement type in the skill requirement tag constitute a tag group. It should be noted that when the tag item corresponding to the job dynamic tag in the tag group is empty, it is set to be consistent with the corresponding tag item in the gig dynamic tag to ensure that the cosine similarity of the tag group is 1; however, if the corresponding tag item in the gig dynamic tag is empty, the previous or default tag item can be used instead.

[0035] For example, the label item group corresponding to the dimension label group is marked as , It is a positive integer; calculated using the formula. Obtain the cosine similarity of the label item groups in each dimension label group. In this formula For tag item group The weighting coefficients can also be set based on the analytic hierarchy process (AHP). By default, the weighting coefficients for each label group are equal. It is the cosine similarity of the label group.

[0036] J03: The basic matching degree is obtained by weighted summation based on the weight coefficient and the tag similarity; the preset matching degree threshold is retrieved, and the person-job matching group is generated based on the basic matching degree and the matching degree threshold.

[0037] After determining the weight coefficients of each dimension of the label group and calculating the label similarity of each label group, a weighted summation can be performed to obtain the basic matching degree.

[0038] For example, the dimension label group is marked as , It is a positive integer; through the formula Calculate the basic matching degree In this formula For dimension label groups The weighting coefficients.

[0039] After calculating the basic matching degree of the dynamic tags for temporary workers and job positions, a preset matching degree threshold is retrieved. This threshold is used to filter out combinations that do not meet the basic matching degree requirements, and then a person-job matching group is generated based on the combinations that meet the requirements. The matching degree threshold is set based on experience and can be dynamically adjusted according to actual conditions; the default matching degree threshold is 0.6.

[0040] In the job matching group, both casual workers and regular jobs correspond to several matching combinations. Due to the volatility of supply and demand for casual workers, it is obviously impossible to match and recommend jobs based on all matching combinations. Therefore, it is necessary to calculate the comprehensive matching score of each matching combination and determine the optimal matching combination for casual workers or regular jobs based on the comprehensive matching score.

[0041] The calculation process for the comprehensive matching score for each matching combination in the job-person matching group is as follows: D01: Set the weights for supply and demand adjustment and labor cost based on the supply and demand imbalance index; In the job matching group, there are several matching combinations for the same gig worker or the same job position. The basic matching degree of these matching combinations all meet the gig worker matching requirements. However, not all matching combinations can be put into practical use. Either a matching combination is selected as the recommended job for gig workers, or the matching combinations are further filtered.

[0042] After obtaining the job-person matching group, this invention sets the supply and demand adjustment weight and the labor cost weight according to the supply and demand imbalance index. If the supply of gig workers in the area where the job is located in the matching group exceeds the demand, the labor cost weight is increased to improve the gig worker experience. When the supply of gig workers is insufficient, the supply and demand adjustment weight is increased to ensure the matching degree between the job and the gig worker.

[0043] D02: Calculate the comprehensive matching score of each matching combination in the job matching group by combining the basic matching degree and labor cost; and achieve matching of casual workers and job positions based on the comprehensive matching score.

[0044] First, extract the basic matching degree of the matching combination in the person-job matching group and calculate the labor cost by combining the temporary work position and the job position. Then, combine the set supply and demand adjustment weight and labor cost weight to calculate the comprehensive matching score of each matching combination. The higher the comprehensive matching score, the higher the person-job fit.

[0045] It should be noted that each job or temporary worker in the job matching group corresponds to several matching combinations. To improve calculation efficiency, the comprehensive evaluation score of the matching combination can be calculated from either the perspective of the temporary worker or the perspective of the company. For example, from the perspective of temporary workers, the matching combinations corresponding to each temporary worker can be extracted according to the upload time or other indicators, and the comprehensive matching score of the matching combination can be calculated sequentially.

[0046] After calculating the comprehensive matching score of each matching combination in the job matching group, the optimal matching combination between temporary workers and job positions is selected based on the comprehensive matching score, and the matching between temporary workers and job positions is realized based on the optimal matching combination.

[0047] For example, suppose there are two job seekers, A and B. Job seeker A uploaded their job search data earlier than job seeker B. The matching combinations for job seeker A include {(A, Job 1), (A, Job 2), (A, Job 3)}, and the matching combinations for job seeker B include {(B, Job 3), (B, Job 4), (B, Job 5)}. Both job seekers A and B have one job posting, and each job posting is intended to hire one job seeker. First, using job seeker A as the base, calculate the comprehensive evaluation score of their matching combinations, and select the matching combination with the highest comprehensive evaluation score as job seeker A's optimal combination. If job seeker A's optimal matching combination is (A, Job 3), since job 3 does not hire job seekers, from job seeker B's perspective, it is only necessary to calculate the comprehensive evaluation score of {(B, Job 4), (B, Job 5)}, and select job seeker B's optimal matching combination from among them.

[0048] It should be noted that if worker A abandons the optimal matching combination (A, position 3), and worker B has not confirmed the optimal matching combination, then position 3 is released to recalculate the comprehensive matching score of the matching combination (B, position 3). The comprehensive matching score is then compared with the comprehensive matching scores of the matching combinations (B, position 4) and (B, position 5) to determine worker B's optimal matching combination.

[0049] When setting the supply and demand adjustment weight and the labor cost weight based on the supply and demand imbalance index, the supply and demand adjustment weight is actually calculated using the supply and demand imbalance index, and the labor cost weight is obtained by using the sum of the supply and demand adjustment weight and the labor cost weight as 1.

[0050] The purpose of supply and demand adjustment weights is to balance the supply and demand of gig workers within a target area. When the supply of gig workers exceeds the demand, it indicates that competition for gig workers in the target sub-area is fierce. The demand for gig workers should be to quickly obtain suitable orders and reduce labor costs. Therefore, when calculating supply and demand adjustment weights using the supply and demand imbalance index, the higher the supply and demand imbalance index, the lower the supply and demand adjustment weight, and vice versa.

[0051] The supply and demand adjustment function is a pre-built function used to calculate the supply and demand adjustment weights based on the supply and demand imbalance index. The supply and demand adjustment weights obtained by this function should not exceed the preset range of the supply and demand adjustment weights. The preset range is set based on experience, such as (0, 1).

[0052] In one example, assume the supply-demand imbalance index is... Then the supply and demand adjustment weight In this formula The preset value for adjusting the weights based on supply and demand. The default value is 0.5. The adjustment coefficient is used to control the range of adjustment for supply and demand adjustment weights. The default value is 0.2; For normalization processing, Indicates will Map to the range [-1, 1].

[0053] After determining the supply and demand adjustment weights and the labor cost weights, the labor cost for each matching combination in the job-person matching group is calculated. This labor cost includes food costs and commuting costs. Next, the basic matching degree of the matching combination is extracted. Based on the supply and demand adjustment weights and the labor cost weights, the basic matching degree and the labor cost are weighted and summed to obtain the comprehensive matching score of the matching combination.

[0054] Labor costs refer to the expenses incurred by temporary workers in performing their duties, primarily including food costs and commuting costs. Most people cannot clearly define their labor costs when doing temporary work, and the acceptable labor costs vary depending on the time of day and the specific job. Therefore, labor costs need to be calculated for each combination of tasks. When calculating labor costs, commuting costs should be estimated based on commuting distance and time, and food costs should be estimated based on commuting time and actual working hours.

[0055] It should be noted that if the job salaries are the same but the labor costs are different, it is not necessarily true that the job with lower labor costs is a better match. It is also necessary to take into account factors such as job type and commuting method. Moreover, the weight of salary expectations in the calculation of basic matching degree may change dynamically. Therefore, it is not advisable to directly add labor costs to salary expectations.

[0056] For example, suppose the supply and demand adjustment weights are... The labor cost weighting is ,and ; can be achieved through formula The overall matching score of the matching combination is calculated. In this formula To match the labor costs corresponding to the combination, A preset cost threshold for gig workers.

[0057] The supply-demand imbalance index is used to calculate the weights for supply and demand adjustments and is the basis for quantifying the suitability of matching groups. The supply-demand imbalance index primarily characterizes the degree of supply-demand imbalance in the gig economy, and its calculation is based on the number of gig workers and the number of job openings. The calculation process for the supply-demand imbalance index is as follows: G01: Retrieve job supply and demand data for the target area; In the intelligent matching process for gig workers, if only the matching degree between gig workers and jobs is considered and the gig worker recommendations are ultimately completed based on this matching degree, it is very likely that supply and demand imbalances will occur in some areas of the target region. To ensure that supply and demand imbalances are avoided during the intelligent matching and recommendation of gig workers, the supply and demand data of gig workers and jobs in the target region are retrieved first. The supply and demand data of gig workers and jobs is obtained by statistical analysis of gig worker job search data and job posting data. The supply and demand data of gig workers and jobs includes specific locations in order to calculate the supply and demand imbalance index of the target sub-region.

[0058] G02: Calculate the supply and demand imbalance index of several target sub-regions in the target region based on the supply and demand data of workers and jobs; among them, the supply and demand imbalance index is used to characterize the degree of supply and demand imbalance in the gig economy.

[0059] After obtaining the supply and demand data for workers and jobs in the target area, a supply-demand imbalance index for several target sub-regions within the target area is calculated based on this data. The supply-demand imbalance index characterizes the supply and demand status of gig workers relative to job positions in the corresponding target sub-region. Specifically, the ratio of the number of gig workers to the number of job positions in the target sub-region can be used as the supply-demand imbalance index. A supply-demand imbalance index of 1 indicates a balance between supply and demand, a supply-demand imbalance index greater than 1 indicates an oversupply of gig workers, and a supply-demand imbalance index less than 1 indicates an undersupply of gig workers.

[0060] It should be noted that the target area includes several target sub-areas. The target area and target sub-areas can be determined using existing administrative divisions or by defining them themselves. For example, a municipal-level administrative region can be used as the target area, and county-level or district-level administrative regions can be used as target sub-areas. It is even possible to use streets or communities as target sub-areas.

[0061] This invention calculates the comprehensive matching score of each matching combination in the job-person matching group based on the supply-demand imbalance index, and then selects the optimal matching combination based on the comprehensive matching score. The purpose is to accurately quantify the suitability priority of each matching combination and solve the problem of how to sort multiple matching combinations. This can ensure the accuracy of matching, take into account the balance of supply and demand and labor costs, improve the willingness of gig workers to accept orders and the efficiency of enterprise employment, and achieve multi-objective optimization of job-person matching.

[0062] Please see Figure 3To balance the supply and demand of gig workers in each target sub-region, the basic matching degree of each matching combination in the job-person matching group is optimized based on the supply-demand imbalance index before calculating the comprehensive matching score of the matching combinations. The specific optimization logic is as follows: if the supply of gig workers in a target sub-region exceeds demand, the basic matching degree associated with that target sub-region is reduced; if the supply of gig workers in a target sub-region falls short of demand, the basic matching degree associated with that target sub-region is increased. The matching degree adjustment ratio for the target sub-region is calculated based on the supply-demand imbalance index, and this adjustment ratio is used to optimize the basic matching degree. If the optimized basic matching degree is less than the matching degree threshold, the matching combination still needs to be removed from the job-person matching group.

[0063] For example, suppose the current supply of logistics gig workers in target sub-region A (city center business district) is 12, and the demand for positions is 20, indicating a supply shortage; while the supply of logistics gig workers in target sub-region B (suburban industrial park) is 30, and the demand for positions is 10, indicating a supply surplus. If gig worker A (logistics skills, currently in region A, basic matching degree 0.8) is matched with both position X (logistics delivery, basic matching degree 0.8) in target sub-region A and position Y (logistics sorting, basic matching degree 0.82) in target sub-region B, then according to the supply and demand optimization rules, the basic matching degree between gig worker A and position X in target sub-region A will be increased by 15% to 0.92, and position X will be recommended to gig worker A first. Meanwhile, the matching priority of position Y in target sub-region B will be reduced, and gig workers in target sub-region B will be matched first, avoiding resource mismatch of vacant positions in target sub-region A and idle gig workers in target sub-region B.

[0064] When adjusting the matching degree of the target sub-region based on the supply and demand imbalance index, the pre-constructed matching degree adjustment function is retrieved, and the supply and demand imbalance index is substituted into the matching degree adjustment function to obtain the matching degree adjustment value.

[0065] The matching degree adjustment function is a function model established with the supply-demand imbalance index as the independent variable and the matching degree adjustment ratio as the dependent variable. A linear function is preferred to construct the matching degree adjustment function. Moreover, the upper and lower limits of the matching degree adjustment ratio are limited, such as the range of the matching degree adjustment ratio being [-0.2, 0.2].

[0066] For example, if the target imbalance index is substituted into the matching degree adjustment function and the matching degree adjustment ratio of the target sub-region is -0.1, then all job-related matching combinations in the target sub-region are extracted, and the basic matching degree in the matching combination is multiplied by (1+(-0.1)) for adjustment, thus completing the optimization of the person-job matching group.

[0067] After generating person-job matching groups based on person-job similarity, some matching combinations show a large commuting distance or long commuting time between the temporary worker and the job, which does not meet the expected cost of employment for temporary workers. These matching combinations should be immediately removed from the subsequent comprehensive matching score calculation to reduce the burden on subsequent data processing. The process for optimizing person-job matching groups based on employment cost is as follows: C01: Calculate the labor cost of each matching combination in the job matching group; As mentioned above, labor costs include food costs and commuting costs; if it is not possible to accurately calculate the labor costs from gig work to a job in the matching combination, the average value of historical data can be used instead.

[0068] C02: Retrieve the preset cost ceiling for gig workers; compare the labor cost with the cost ceiling to delete unsuitable matching combinations in order to optimize the job-person matching group.

[0069] The cost cap is set by gig workers themselves when uploading their job search data, taking into account both food and commuting costs. If the total cost of a matched job exceeds the cost cap, it indicates that while the job in that match is a good fit for gig workers, it exceeds their expected costs. In this case, the match should be removed to optimize the job-person matching system.

[0070] It should be noted that optimizing the job-person matching group based on the supply-demand imbalance index aims to initially balance the supply and demand relationship between temporary workers and regular jobs. If the optimized basic matching degree exceeds its preset constraints (such as exceeding the range of [0, 1]), then the limit value of the preset constraints will be reached. Optimizing the job-person matching group based on labor costs aims to eliminate unreasonable matching combinations from multiple matching combinations. However, if there are not many matching combinations to begin with, such as a temporary worker being associated with only one matching combination, then that matching combination will be retained and sent to the temporary worker for them to decide for themselves. This can improve the success rate of matching temporary workers and regular jobs.

[0071] This invention optimizes the job-person matching group based on the supply-demand imbalance index and labor costs before calculating the comprehensive matching score. Its core function is to eliminate unreasonable matching combinations to screen effective matching combinations, initially balance the supply and demand of the target sub-region, provide high-quality samples for calculating the comprehensive matching score based on the supply-demand imbalance index, and avoid invalid calculations.

[0072] It should be understood that the descriptions of technical features, technical solutions, beneficial effects, or similar language in this application do not imply that all features and advantages can be achieved in any single embodiment. Rather, it is understood that the description of a feature or beneficial effect means that a specific technical feature, technical solution, or beneficial effect is included in at least one embodiment. Therefore, the descriptions of technical features, technical solutions, or beneficial effects in this specification do not necessarily refer to the same embodiment. Furthermore, the technical features, technical solutions, and beneficial effects described in this embodiment can be combined in any suitable manner. Those skilled in the art will understand that embodiments can be implemented without one or more specific technical features, technical solutions, or beneficial effects of a particular embodiment. In other embodiments, additional technical features and beneficial effects may be identified in specific embodiments that do not embody all embodiments.

[0073] In the above embodiments, implementation can be achieved entirely or partially through software, hardware, firmware, or any other combination thereof. When implemented using a software program, it can be implemented entirely or partially in the form of a computer program product. This computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium accessible to a computer or a data storage device containing one or more servers, data centers, etc., that can be integrated with the medium. The available media can be magnetic media (e.g., floppy disks, hard disks, magnetic tapes), optical media (e.g., DVDs), or semiconductor media (e.g., solid-state disks (SSDs)).

[0074] The above embodiments are only used to illustrate the technical methods of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical methods of the present invention without departing from the spirit and scope of the technical methods of the present invention.

Claims

1. A zero-worker job intelligent matching and recommendation system based on a multi-dimensional dynamic tagging system, characterized in that: Includes a job matching module and a tag building module; Tag building module: used to build multi-dimensional dynamic tags; among which, multi-dimensional dynamic tags include gig dynamic tags and job dynamic tags, as well as the dimensional correspondence between gig dynamic tags and job dynamic tags; The job matching module is used to perform job similarity matching based on the multi-dimensional dynamic tags to obtain job matching groups; calculate the comprehensive matching score of each matching combination in the job matching group; and select the optimal matching combination based on the comprehensive matching score to achieve matching between casual workers and job positions.

2. The intelligent matching and recommendation system for zero-worker positions based on a multi-dimensional dynamic tagging system according to claim 1, characterized in that, The dynamic tags for gig workers include basic attribute tags, skill attribute tags, behavioral performance tags, scenario adaptation tags, salary preference tags, and real-time status tags; the dynamic tags for job positions include basic requirement tags, skill requirement tags, performance requirement tags, scenario attribute tags, salary and benefits tags, and real-time requirement tags.

3. The intelligent matching and recommendation system for zero-worker positions based on a multi-dimensional dynamic tagging system according to claim 1, characterized in that, Based on the aforementioned multi-dimensional dynamic tags, person-job similarity matching is performed, including: Based on the dimensional correspondence of the multi-dimensional dynamic labels, dimension label groups are constructed, and weight coefficients are set for each dimension label group. Calculate the label similarity of each of the aforementioned dimension label groups; perform a weighted sum based on the weight coefficients and the label similarity to obtain the basic matching degree; wherein, the label similarity is calculated using cosine similarity. A preset matching threshold is retrieved, and a person-job matching group is generated based on the basic matching degree and the matching threshold; wherein, the default value of the matching threshold is 0.

6.

4. The intelligent matching and recommendation system for zero-worker positions based on a multi-dimensional dynamic tagging system according to claim 1, characterized in that, Calculate the comprehensive matching score for each matching combination in the job matching group, including: The supply and demand imbalance index is used to set weights for supply and demand adjustment and labor cost; the supply and demand imbalance index is used to characterize the degree of supply and demand imbalance in the gig economy. The comprehensive matching score of each matching combination in the job matching group is calculated based on the supply and demand adjustment weight and the labor cost weight; wherein, the comprehensive matching score is used to quantify the adaptation priority of the matching combination.

5. The intelligent matching and recommendation system for zero-worker positions based on a multi-dimensional dynamic tagging system according to claim 4, characterized in that, The supply and demand imbalance index is used to set weights for supply and demand adjustment and labor cost, including: The supply-demand imbalance index is substituted into a pre-constructed supply-demand adjustment function to obtain the supply-demand adjustment weight; wherein, the higher the supply-demand imbalance index, the lower the corresponding supply-demand adjustment weight. The labor cost weight is obtained by solving the problem using the sum of the supply and demand adjustment weight and the labor cost weight, which is 1.

6. The intelligent matching and recommendation system for zero-worker positions based on a multi-dimensional dynamic tagging system according to claim 4, characterized in that, The comprehensive matching score for each matching combination in the job-person matching group is calculated based on the supply-demand adjustment weight and the labor cost weight, including: Calculate the labor cost of the matched combination; wherein, the labor cost includes food cost and commuting cost; Based on the supply and demand adjustment weights and the labor cost weights, the basic matching degree and labor cost of the matching combination are weighted and summed to obtain a comprehensive matching score.

7. The intelligent matching and recommendation system for zero-worker positions based on a multi-dimensional dynamic tagging system according to claim 4, characterized in that, The supply-demand imbalance index is calculated based on the supply and demand data for jobs, including: Retrieve job and labor supply and demand data for the target area; the target area is the region that needs matching and recommendation of gig workers, and the job and labor supply and demand data includes the number of gig workers and positions in the target area; The supply-demand imbalance index of several target sub-regions in the target region is calculated based on the supply and demand data of personnel and positions; wherein, the supply-demand imbalance index is the ratio of the number of casual workers to the number of job positions in the target sub-region.

8. The intelligent matching and recommendation system for zero-worker positions based on a multi-dimensional dynamic tagging system according to claim 4, characterized in that, Optimizing the job-person matching group based on the supply-demand imbalance index includes: The matching degree adjustment ratio of the target sub-region is based on the supply and demand imbalance index; wherein the preset range of the matching degree adjustment ratio is [-0.2, 0.2]; The basic matching degree of the matching combination associated with the job in the target sub-region is adjusted based on the matching degree adjustment ratio to optimize the person-job matching group.

9. The intelligent matching and recommendation system for zero-worker positions based on a multi-dimensional dynamic tagging system according to claim 8, characterized in that, The matching degree adjustment ratio based on the supply and demand imbalance index for the target sub-region includes: Retrieve the pre-constructed matching degree adjustment function; wherein, the matching degree adjustment function is a linear function; By substituting the supply-demand imbalance index of the target sub-region into the matching degree adjustment function, the matching degree adjustment ratio of the target sub-region is obtained.

10. The intelligent matching and recommendation system for zero-worker positions based on a multi-dimensional dynamic tagging system according to claim 1, characterized in that, Optimizing the job-person matching group based on labor costs includes: Calculate the labor cost of each matching combination in the job matching group; Retrieve the preset cost limit for gig workers; compare the labor cost with the cost limit to delete unsuitable matching combinations, thereby optimizing the job-person matching group.

Images