A distribution system and process for mobile crowdsourcing tasks

Abstract

The invention relates to the technical field of mobile crowdsourcing task distribution, and discloses a mobile crowdsourcing task distribution system and process, comprising: recording mobile crowdsourcing tasks that have not been received and executed as first-class tasks, and recording mobile crowdsourcing tasks that have been received and executed as second-class tasks; arbitrarily selecting a first-class task and recording it as a target first-class task; recording the publisher of the target first-class task as a target publisher; establishing a set A; extracting keywords of the task content description of the target first-class task and listing them in the set A; retrieving executors who are qualified to execute the target first-class task in the execution area and recording them as first-class executors; obtaining the task content and evaluation content of the second-class task published by the target publisher to determine the evaluation preference type of the target publisher, determining the execution level of the first-class executor, and pushing the target first-class task to the first-class executor; the scheme effectively improves the task completion rate and user satisfaction.

Description

Technical Field

[0001] The present invention relates to the technical field of distribution process of mobile crowdsourcing tasks, and in particular to a distribution system and process of mobile crowdsourcing tasks. Background Art

[0002] With the development of mobile Internet, the types and number of mobile crowdsourcing tasks are increasing. The publishers of mobile crowdsourcing tasks publish tasks through the platform, and the executors receive and complete tasks through mobile devices, realizing service sharing and efficient distribution; however, in the face of a variety of task requirements, the platform urgently needs an effective distribution process to improve the efficiency of task distribution and meet the differentiated needs of different tasks.

[0003] The traditional mobile crowdsourcing task distribution process usually does not analyze the specific keywords of the task content and cannot effectively identify the core requirements of the task. It usually only relies on simple matching based on region or task type, resulting in a strong blindness in distribution. This lack of matching accuracy directly affects the success rate of the task and the experience of the task publisher, which may lead to a decline in user trust and the loss of platform users.

[0004] The traditional mobile crowdsourcing task distribution process often lacks a sophisticated and accurate matching mechanism, making it difficult to ensure an accurate match between the executors' capabilities and task requirements. Due to insufficient consideration of the executors' capabilities, many tasks are assigned to executors who do not have the corresponding capabilities, resulting in low quality of task completion. Many highly professional mobile crowdsourcing tasks are therefore unable to be assigned to executors with corresponding capabilities, affecting the completion effect of mobile crowdsourcing tasks and the professionalism of the platform.

[0005] In the existing mobile crowdsourcing task distribution process, after the executor receives and executes the mobile crowdsourcing task published by the publisher, the publisher will score the executor. However, different publishers have different scoring standards. For example, if the same executor executes the same type of mobile crowdsourcing tasks twice but published by different publishers within a short period of time, the scores received may be different or even quite different. The existing mobile crowdsourcing task platform often pushes mobile crowdsourcing tasks to the executors based on the scores of the executors, but simply uses the scores and ignores the problem that different publishers have different scoring standards. Therefore, it is impossible to obtain the execution level of each executor, nor the evaluation preference type of each publisher, resulting in the lack of intelligence and accuracy in the distribution of mobile crowdsourcing tasks.

[0006] The existing mobile crowdsourcing task distribution process usually does not consider the task requirements of mobile crowdsourcing tasks, and often prioritizes pushing mobile crowdsourcing tasks to executors with high execution capabilities. However, some mobile crowdsourcing tasks have lower task requirements and do not require executors with higher execution levels. Therefore, this push mechanism ignores executors with lower execution levels, resulting in unbalanced task resources. Summary of the invention

[0007] The present invention provides a mobile crowdsourcing task distribution system and process, which are used to promote the solution of the problems mentioned in the above background technology.

[0008] The present invention provides the following technical solution: Optionally, a distribution process of a mobile crowdsourcing task is characterized by: comprising:

[0009] According to the publisher's evaluation preference type and the executor's execution level, the mobile crowdsourcing tasks published by the publisher are pushed to the executor, specifically:

[0010] The mobile crowdsourcing tasks that have not been received and executed are recorded as the first-class tasks, and the mobile crowdsourcing tasks that have been received and executed are recorded as the second-class tasks;

[0011] Randomly select a class-one task and record it as the target class-one task;

[0012] The publisher of the target type 1 task is recorded as the target publisher;

[0013] Get the task content of the target type task;

[0014] The task content includes the task location and task content description;

[0015] Establish a keyword set, recorded as set A;

[0016] Extract the keywords of the task content description of the target type one task and list the keywords in set A;

[0017] Determine the executors who are qualified to perform the target type of tasks based on the keywords in set A;

[0018] Retrieve the executors who are qualified to execute the target type one task in the execution area and record them as type one executors;

[0019] Obtain the task content and evaluation content of the second-category task published by the target publisher;

[0020] Determine the evaluation preference type of the target publisher according to the task content and evaluation content of the second type of tasks published by the target publisher;

[0021] Obtain the task execution experience of the first-class executors, and determine the execution level of the first-class executors in combination with the evaluation content of the second-class tasks published by the target publisher;

[0022] According to the evaluation preference type of the target publisher and the execution level of the first type of executors, a target type of tasks is pushed to the first type of executors.

[0023] Optionally, the executors who are qualified to execute the target type one task in the search execution area and record them as type one executors are specifically:

[0024] The area selected by the target publisher when publishing the target type task is recorded as the execution area;

[0025] Obtain the task execution experience of all executors in the execution area;

[0026] The task execution experience includes the task content of the task executed and the evaluation level made by the publisher after the task is executed;

[0027] The evaluation level is specifically 0, 100, and a whole number within 100, wherein the larger the number, the higher the evaluation level;

[0028] Set the keyword proportion threshold Z1;

[0029] S1. Randomly select an executor in the execution area;

[0030] S2. Obtain the task content description of all task execution experiences of this executor;

[0031] S3, extracting keywords describing the task content in all the task execution experiences of this executor;

[0032] S4, establish a keyword set, record it as set B, and put the extracted keywords into set B;

[0033] S5. Randomly select a keyword in set A. If the selected keyword in set A exists in set B, the selected keyword in set A is recorded as a matching keyword;

[0034] S6. Obtain the number of keywords and matching keywords in set A respectively, record them as A1 and A2 respectively, and calculate the ratio of A2 to A1, record them as A2÷A1;

[0035] S7. If A2÷A1≥Z1, then this executor is qualified to receive and execute the target type 1 task;

[0036] S8. If A2÷A1<Z1, then this executor is not qualified to receive and execute the target type 1 task;

[0037] Perform S1-S8 operation steps on all executors in the execution area to determine all executors in the execution area who are qualified to execute the target type one task, and record all executors in the execution area who are qualified to execute the target type one task as type one executors.

[0038] Optionally, the evaluation preference type of the target publisher is determined according to the task content and evaluation content of the second type of task published by the target publisher, specifically:

[0039] Get the number of executors who have executed the second-class tasks issued by the target publisher in the past thirty days, and record the number as C;

[0040] The C executors who execute the second-category tasks issued by the target publisher within nearly thirty days are respectively recorded as the first executor, the second executor, the third executor, and the Cth executor;

[0041] Obtain the task execution experience of the first executor, the second executor, the third executor, ... the Cth executor in sequence;

[0042] S9. Select any one executor from the first executor, the second executor, the third executor, ... the Cth executor:

[0043] S10, obtaining the number of the second-category tasks executed by the executor in the past thirty days, the number being recorded as D;

[0044] S11, respectively obtain the evaluation level of the executor after the executor performs D second-class tasks, and the evaluation levels are recorded as D1, D2, D3...D D ;

[0045] S12, obtaining the evaluation level made by the target publisher after the executor performs the second type of task of the target publisher, where the evaluation level is recorded as d;

[0046] S13, obtain D1, D2, D3...D respectively D The absolute value of the difference from d;

[0047] S14, setting a difference threshold Z2, and setting an evaluation level proportion threshold Z3;

[0048] S15, if D1, D2, D3...D D If the ratio of the number of absolute values ​​of the difference between the values ​​of Z2 and d is greater than or equal to Z3, the evaluation type of the target publisher on this executor is recorded as a first-class evaluation;

[0049] S16, if D1, D2, D3...D D Among the absolute values ​​of the differences from d, if the ratio of the number of absolute values ​​less than or equal to Z2 to D is less than Z3, then D1, D2, D3, ..., D are obtained respectively.D The number of evaluation levels with values ​​greater than d and less than d are recorded as E and F respectively;

[0050] S17. If E≥F, the target publisher's evaluation type for the executor is recorded as a second-class evaluation; if E<F, the target publisher's evaluation type for the executor is recorded as a third-class evaluation;

[0051] Performing operations S9-S17 on the first executor, the second executor, the third executor, ... the Cth executor in sequence, obtaining evaluation types of the target publisher on the first executor, the second executor, the third executor, ... the Cth executor;

[0052] The task execution experience of the first-class executors is obtained, and the execution level of the first-class executors is determined in combination with the evaluation content of the second-class tasks published by the target publisher.

[0053] Optionally, the step of determining the evaluation preference type of the target publisher according to the task content and evaluation content of the second type of task published by the target publisher further includes:

[0054] Obtain the number of first-category evaluations, second-category evaluations, and third-category evaluations of the C evaluation types of the target publisher on the first executor, the second executor, the third executor, ..., the Cth executor, respectively, and record the numbers as G, H, and J respectively;

[0055] If G ≥ H and G > J, or G > H and G ≥ J, the target publisher's evaluation preference type is recorded as a type of evaluation preference;

[0056] If H>G and H≥J, the target publisher’s evaluation preference type is recorded as the second-class evaluation preference;

[0057] If J>G and J>H, the target publisher’s evaluation preference type is recorded as three types of evaluation preference;

[0058] If G=H=J, the evaluation preference types of the target publisher are recorded as four types of evaluation preferences.

[0059] Optionally, the task execution experience of the first type of executor is obtained, and the execution level of the first type of executor is determined in combination with the evaluation content of the second type of task published by the target publisher, specifically:

[0060] Get the evaluation level of the target publisher on the first executor, the second executor, the third executor, ... the Cth executor, which are recorded as C1, C2, C3, ... C respectively. C ;

[0061] Get C1, C2, C3, ... C C The average value is recorded as L, L = (C1 + C2 + C3 + ... + C C )÷C;

[0062] S18. Randomly select a first-class executor and obtain the number of second-class tasks executed by this type of executor in the past thirty days, and record the number as K;

[0063] S19, obtain K evaluation levels made by the publisher to this type of executor after the executor of this type performs K second-type tasks, and record them as K1, K2, K3...K K ;

[0064] S20, calculate the average value of K evaluation levels, the average value is recorded as V, V = (K1 + K2 + K3 + ... + K K )÷K;

[0065] S21. If the difference between V and L VL>Z2, then this type of performer is recorded as a high-level performer. If the difference between V and L LV>Z2, then this type of performer is recorded as a low-level performer. If the absolute value of the difference between V and L |VL|≤Z2, then this type of performer is recorded as a medium performer.

[0066] Perform S18-S21 operations on all executors of the first category.

[0067] Optionally, the method of pushing a target type of tasks to a type of executors according to the evaluation preference type of the target publisher and the evaluation type of a type of executors is specifically as follows:

[0068] If the evaluation preference type of the target publisher is a type of evaluation preference, search for medium performers in the execution area;

[0069] If there are intermediate executors in the execution area, the task push strategy is executed on the intermediate executors, specifically:

[0070] Get the distance between the location of each executor and the task location of the target task;

[0071] According to the distance between the location of each executor and the task location of the target type 1 task, the target type 1 tasks are pushed to the executors in order from near to far;

[0072] If there is no middle executor in the execution area, then search for high-level executors in the execution area. If there is a high-level executor in the execution area, then push the task strategy to the high-level executor.

[0073] If neither the high-level executor nor the medium-level executor exists in the execution area, the target type 1 task will be put into the waiting queue to wait until a medium-level executor or a high-level executor exists in the execution area;

[0074] If the evaluation preference type of the target publisher is the second-class evaluation preference, search for high-level performers in the execution area;

[0075] If there is a high-level executor in the execution area, the task push strategy is executed on the high-level executor;

[0076] If there is no high-level executor in the execution area, the target type 1 task will be placed in the waiting queue to wait until there is a high-level executor in the execution area.

[0077] Optionally, the step of pushing a target type of tasks to a type of executors according to the evaluation preference type of the target publisher and the evaluation type of a type of executors further includes:

[0078] If the target publisher’s evaluation preference type is recorded as three types of evaluation preference, low-level performers are retrieved in the execution area;

[0079] If there are low-level executors in the execution area, the task push strategy is executed on the low-level executors;

[0080] If there is no low-level executor in the execution area, then search for a medium-level executor in the execution area. If there is a medium-level executor in the execution area, then execute the task push strategy to the medium-level executor.

[0081] If neither the low-level executor nor the middle-level executor exists in the execution area, the high-level executor is searched in the execution area. If the high-level executor exists in the execution area, the task push strategy is executed on the high-level executor.

[0082] If there are no low-level executors, medium-level executors, or high-level executors in the execution area, the target type 1 task will be put into the waiting queue to wait until there are low-level executors and low-level executors, medium-level executors, or high-level executors in the execution area;

[0083] If the evaluation preference type of the target publisher is the four-category evaluation preference, search for high-level performers in the execution area;

[0084] If there is a high-level executor in the execution area, the task push strategy is executed on the high-level executor;

[0085] If there is no high-level executor in the execution area, then search for medium-level executors in the execution area. If there is a medium-level executor in the execution area, then execute the task push strategy to the medium-level executor.

[0086] If neither the high-level executor nor the middle-level executor exists in the execution area, the low-level executor is searched in the execution area. If the low-level executor exists in the execution area, the task push strategy is executed to the high-level executor.

[0087] If there are no low-level executors, medium-level executors, or high-level executors in the execution area, the target type 1 task will be placed in a waiting queue and wait until there are low-level executors and low-level executors, medium-level executors, or high-level executors in the execution area.

[0088] Optionally, a system for implementing the distribution process of a mobile crowdsourcing task is characterized by comprising:

[0089] Task content extraction module: obtains the task content of mobile crowdsourcing tasks and extracts keywords describing the task content of mobile crowdsourcing tasks;

[0090] Matching module: determines the executors who are qualified to perform the mobile crowdsourcing tasks in the execution area based on keywords;

[0091] Data processing module: determines the publisher's evaluation preference type and the executor's execution level based on the task content and evaluation content of the second-category task published by the publisher;

[0092] Task push module: pushes mobile crowdsourcing tasks to executors based on the publisher's evaluation preference type and the executor's execution level.

[0093] The present invention has the following beneficial effects:

[0094] 1. The distribution process of the mobile crowdsourcing task is to extract the keywords of the task content description of the target type one task, so that the mobile crowdsourcing task platform can quickly identify the task requirements, and then obtain the task execution experience of each executor in the execution area, and obtain the executor's ability field through the keywords of the task content description in the task execution experience, and compare and screen the publisher's task requirements with the executor's ability field; establish a keyword set, and retrieve the executors who are qualified to execute the target type one tasks in the execution area according to the keywords in the keyword set. This solution greatly improves the matching accuracy of the target type one tasks and executors, ensures that the target type one tasks are distributed to executors with relevant capabilities, improves the quality of task completion, and reduces unnecessary task allocation errors and the need for reallocation.

[0095] 2. The distribution process of the mobile crowdsourcing task, when obtaining the evaluation preference type of the target publisher, the C executors who executed the second-category tasks of the target publisher in the past thirty days are respectively recorded as the first executor, the second executor, the third executor...the Cth executor. Randomly select one of the first executor, the second executor, the third executor...the Cth executor, first obtain the evaluation level of the target publisher on this executor, and then obtain the D evaluation levels corresponding to the D task execution experiences of this executor in the past thirty days. If the evaluation levels of the other D publishers on this executor are generally higher than the evaluation level of the target publisher on this executor, it means that the target publisher has higher requirements for the task execution standards than the other D publishers. Therefore, the evaluation of the target publisher on this executor is recorded as the second-category evaluation; if the evaluation levels of the other D publishers on this executor are generally higher than the evaluation level of the target publisher on this executor, it means that the target publisher has higher requirements for the task execution standards than the other D publishers. Therefore, the evaluation of the target publisher on this executor is recorded as the second-category evaluation; If the evaluation level of the target publisher on this executor is generally lower than that of the target publisher on this executor, it means that the target publisher's requirements for the task execution standards are generally lower than those of the other D publishers. Therefore, the target publisher's evaluation of this executor is recorded as a third-category evaluation; if the evaluation levels of the other D publishers on this executor are roughly at the same level as the evaluation level of the target publisher on this executor, it means that the target publisher's requirements for the task execution standards are roughly the same as those of the other D publishers. Therefore, the target publisher's evaluation of this executor is recorded as a first-category evaluation; through this method, the platform can obtain the target publisher's evaluation type of the first executor, the second executor, the third executor...the Cth executor, and then obtain the target publisher's evaluation preference type, and screen out executors who are more in line with the publisher's evaluation preference type, which can not only improve the accuracy of task completion, but also improve the publisher's satisfaction.

[0096] 3. The distribution process of the mobile crowdsourcing task, after obtaining the target publisher's evaluation types for the first executor, the second executor, the third executor ... the Cth executor, calculate the proportion of the first, second and third evaluation types in the C evaluation types; if the proportion of the first evaluation type is large, it means that the majority of the C evaluation levels made by the target publisher are consistent with the executor's execution ability, so the evaluation level made by the target publisher is neither too high nor too low, and the target publisher's evaluation preference type is recorded as the first evaluation preference type; if the proportion of the second evaluation type is large, it means that the minority of the C evaluation levels made by the target publisher are consistent with the executor's execution ability, and most of the evaluation levels made are lower than the executor's execution ability, so the evaluation level made by the target publisher is low, and the target publisher's evaluation preference type is recorded as the second evaluation preference type; If the proportion of one evaluation, two evaluations and three evaluations is equal, it means that there is no fixed standard for the target publisher when making evaluations, so the target publisher's evaluation preference type is recorded as four evaluation preference types; this classification mechanism can standardize the evaluation preferences of different publishers, so that the evaluation levels obtained by each executor are unified and standardized, which is more valuable for reference. For publishers, they can get the service experience of executors that meet their preferences, which makes them more confident in subsequent use, thereby increasing the stickiness of the platform.

[0097] 4. The distribution process of the mobile crowdsourcing task is to divide the first-class executors into high-level executors, medium-level executors and low-level executors by calculating the average value of the evaluation level of each first-class executor and then comparing the average value with the average value of the evaluation level of the first executor, the second executor, the third executor ... the Cth executor by the target publisher; when distributing tasks, if the evaluation preference type of the target publisher is the first-class evaluation preference type, the task is distributed to the medium-level executor first; if the evaluation preference type of the target publisher is the second-class evaluation preference type or the fourth-class evaluation preference type, the task is distributed to the high-level executor first and only; if the evaluation preference type of the target publisher is the third-class evaluation preference type, the task is distributed to the low-level executor first; such a grading system can not only help the platform make more accurate decisions when distributing tasks, but also motivate executors of different execution levels to improve their own levels by completing high-quality tasks; it can also ensure that task resources are more balanced and that executors of different execution levels have the same opportunities when accepting tasks.

[0098] 5. The distribution process of the mobile crowdsourcing task ensures that tasks are distributed preferentially to executors near the task location for executors with the same execution level in the execution area. This strategy significantly improves the response speed and execution efficiency of task distribution, and is especially suitable for geographically sensitive tasks, such as on-site data collection and regional services. When a task cannot find a qualified executor in a short period of time, the platform will add the task to the waiting queue so that it can be distributed after a qualified executor appears, ensuring that high-quality tasks can be distributed to suitable executors. This mechanism improves the distribution patience of the platform, so that important or high-quality tasks are more guaranteed to be completed, and also helps to improve the overall task completion rate of the platform and enhance user trust. BRIEF DESCRIPTION OF THE DRAWINGS

[0099] Figure 1 It is a schematic diagram of the structure of the present invention. DETAILED DESCRIPTION

[0100] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0101] Embodiment 1, a distribution process of a mobile crowdsourcing task, characterized by comprising:

[0102] According to the publisher's evaluation preference type and the executor's execution level, the mobile crowdsourcing tasks published by the publisher are pushed to the executor, specifically:

[0103] The mobile crowdsourcing tasks that have not been received and executed are recorded as the first-class tasks, and the mobile crowdsourcing tasks that have been received and executed are recorded as the second-class tasks;

[0104] Randomly select a class-one task and record it as the target class-one task;

[0105] The publisher of the target type 1 task is recorded as the target publisher;

[0106] Get the task content of the target type task;

[0107] The task content includes the task location and task content description;

[0108] Establish a keyword set, recorded as set A;

[0109] Extract the keywords of the task content description of the target type one task and list the keywords in set A;

[0110] According to the keywords in set A, the executors who are qualified to perform the target type of tasks are determined. In the mobile crowdsourcing scenario, the types of tasks are diverse and may involve different professional fields, such as questionnaires, on-site photography, data collection, etc. Therefore, the keywords of different tasks are also different;

[0111] Retrieve the executors who are qualified to execute the target type one task in the execution area and record them as type one executors;

[0112] Obtain the task content and evaluation content of the second-category task published by the target publisher;

[0113] Determine the evaluation preference type of the target publisher according to the task content and evaluation content of the second type of tasks published by the target publisher;

[0114] Obtain the task execution experience of the first-class executors, and determine the execution level of the first-class executors in combination with the evaluation content of the second-class tasks published by the target publisher;

[0115] According to the evaluation preference type of the target publisher and the execution level of the first type of executors, a target type of tasks is pushed to the first type of executors.

[0116] The executors who have the qualification to execute the target type one task in the search execution area are recorded as type one executors, specifically:

[0117] The area selected by the target publisher when publishing the target type task is recorded as the execution area;

[0118] Obtain the task execution experience of all executors in the execution area;

[0119] The task execution experience includes the task content of the task executed and the evaluation level made by the publisher after the task is executed;

[0120] The evaluation level is specifically 0, 100, and a whole number within 100, wherein the larger the number, the higher the evaluation level;

[0121] Set the keyword ratio threshold Z1. The function of Z1 is to judge whether the executor is qualified to perform the target type 1 task according to the ratio of the matching keywords in set A. The matching keywords represent that the executor's ability matches the requirements of the target type 1 task. When the ratio of the matching keywords is greater than or equal to Z1, it means that the executor is qualified to perform the target type 1 task. When the ratio of the matching keywords is less than Z1, it means that the executor is not qualified to perform the target type 1 task.

[0122] S1. Randomly select an executor in the execution area;

[0123] S2. Obtain the task content description of all task execution experiences of this executor;

[0124] S3, extracting keywords from the task content description in all the task execution experiences of the executor. The task execution experience represents the tasks that the executor has performed. Therefore, the keywords from the task content description in the task execution experience represent the areas that the executor is good at. For example, the executor has performed tasks such as photo processing, video editing, and event shooting. Therefore, the keywords from the task content description in all the task execution experiences of the executor will include keywords such as photography, photo editing, music matching, color grading, video, and editing.

[0125] S4, establish a keyword set, record it as set B, and put the extracted keywords into set B;

[0126] S5. Randomly select a keyword in set A. If the selected keyword in set A exists in set B, the selected keyword in set A is recorded as a matching keyword. For example, the task posted by the target publisher is video editing, and the task content is described as "editing several clips into a 15-second video and matching it with music". Therefore, the keywords in set A are "short video, editing and music matching", and the matching keyword is "editing and music matching".

[0127] S6. Obtain the number of keywords and matching keywords in set A respectively, record them as A1 and A2 respectively, and calculate the ratio of A2 to A1, record them as A2÷A1;

[0128] S7. If A2÷A1≥Z1, then this executor is qualified to receive and execute the target type 1 task;

[0129] S8. If A2÷A1<Z1, then this executor is not qualified to receive and execute the target type 1 task;

[0130] Perform S1-S8 operation steps on all executors in the execution area to determine all executors in the execution area who are qualified to execute the target type one task, and record all executors in the execution area who are qualified to execute the target type one task as type one executors.

[0131] The evaluation preference type of the target publisher is determined according to the task content and evaluation content of the second type of tasks published by the target publisher, specifically:

[0132] Get the number of executors who have executed the second-class tasks issued by the target publisher in the past thirty days, and record the number as C;

[0133] The C executors who execute the second-category tasks issued by the target publisher within nearly thirty days are respectively recorded as the first executor, the second executor, the third executor, and the Cth executor;

[0134] Obtain the task execution experience of the first executor, the second executor, the third executor, ... the Cth executor in sequence;

[0135] S9. Select any one executor from the first executor, the second executor, the third executor, ... the Cth executor:

[0136] S10, obtaining the number of the second-category tasks executed by the executor in the past thirty days, the number being recorded as D;

[0137] S11, respectively obtain the evaluation level of the executor after the executor performs D second-class tasks, and the evaluation levels are recorded as D1, D2, D3...D D ;

[0138] S12, obtaining the evaluation level made by the target publisher after the executor performs the second type of task of the target publisher, where the evaluation level is recorded as d;

[0139] S13, obtain D1, D2, D3...D respectively D The absolute value of the difference from d, which reflects D1, D2, D3...D D degree of difference from d;

[0140] S14, set the difference threshold Z2, the role of Z2 is to judge D1, D2, D3 ... D by comparing the absolute value of each value with Z2. D The difference from d, when the absolute value is greater than Z2, indicates that D1, D2, D3...D D The difference from d is large. When the absolute value is less than or equal to Z2, it means that D1, D2, D3...D D The difference with d is small, and the evaluation level ratio threshold Z3 is set. The role of Z3 is to judge the evaluation difference between the target publisher and the other D publishers through the ratio of absolute values;

[0141] S15, if D1, D2, D3...D D The ratio of the number of absolute values ​​of the difference with d that is less than or equal to Z2 to D is greater than or equal to Z3. The absolute value is less than or equal to Z2, which means that D1, D2, D3...D D The difference from d is small, and the ratio is greater than or equal to Z3, which means that D1, D2, D3...D D The absolute values ​​with smaller differences from d account for the majority, that is, the evaluation level of the other D publishers on this executor is roughly at the same level as the evaluation level of the target publisher on this executor, indicating that the target publisher's requirements for the task execution standard are roughly the same as those of the other D publishers. In this case, the evaluation type of the target publisher on this executor is recorded as a first-class evaluation;

[0142] S16, if D1, D2, D3...D DThe ratio of the number of absolute values ​​of the difference between Z1 and d that is less than or equal to Z2 to D is less than Z3. If the ratio is less than Z3, it means that D1, D2, D3, ... D D The absolute values ​​with a large difference from d account for the majority, and it is necessary to further determine the evaluation type of the target publisher on this executor, and then obtain D1, D2, D3...D D The number of evaluation levels with values ​​greater than d and less than d are recorded as E and F respectively;

[0143] S17. If E≥F, that is, the evaluation level of the other D publishers on this executor is generally higher than the evaluation level of the target publisher on this executor, indicating that the target publisher's requirements for the task execution standard are generally higher than those of the other D publishers, then the target publisher's evaluation type for this executor is recorded as a second-class evaluation; if E<F, that is, the evaluation level of the other D publishers on this executor is generally lower than the evaluation level of the target publisher on this executor, indicating that the target publisher's requirements for the task execution standard are generally lower than those of the other D publishers, then the target publisher's evaluation type for this executor is recorded as a third-class evaluation;

[0144] Performing operations S9-S17 on the first executor, the second executor, the third executor, ... the Cth executor in sequence, obtaining evaluation types of the target publisher on the first executor, the second executor, the third executor, ... the Cth executor;

[0145] The task execution experience of the first type of executor is obtained, and the execution level of the first type of executor is determined in combination with the evaluation content of the second type of task published by the target publisher. Through this method, the platform can obtain the evaluation type of the target publisher on the first executor, the second executor, the third executor...the Cth executor, and then obtain the evaluation preference type of the target publisher, and screen out executors who are more in line with the publisher's evaluation preference type, which can not only improve the accuracy of task completion, but also improve the publisher's satisfaction.

[0146] The step of determining the evaluation preference type of the target publisher according to the task content and evaluation content of the second type of task published by the target publisher also includes:

[0147] Obtain the number of first-category evaluations, second-category evaluations, and third-category evaluations of the C evaluation types of the target publisher on the first executor, the second executor, the third executor, ..., the Cth executor, respectively, and record the numbers as G, H, and J respectively;

[0148] If G≥H and G>J, or G>H and G≥J, that is, the proportion of the first-class evaluation is relatively large, indicating that the evaluation levels that match the execution ability of the executor account for the majority of the C evaluation levels made by the target publisher. Therefore, the evaluation level made by the target publisher is neither too high nor too low. The evaluation preference type of the target publisher is recorded as the first-class evaluation preference;

[0149] If H>G and H≥J, that is, the proportion of the second-class evaluation is relatively large, it means that the evaluation levels that match the execution ability of the executor account for a small number of the C evaluation levels made by the target publisher, and most of the evaluation levels made are lower than the execution ability of the executor. Therefore, the evaluation level made by the target publisher is low, and the evaluation preference type of the target publisher is recorded as the second-class evaluation preference;

[0150] If J>G and J>H, that is, the proportion of the three types of evaluations is relatively large, it means that the evaluation levels that match the execution ability of the executor account for a small number of the C evaluation levels made by the target publisher, and most of the evaluation levels made are higher than the execution ability of the executor. Therefore, the evaluation level made by the target publisher is relatively high, and the evaluation preference type of the target publisher is recorded as the three-type evaluation preference;

[0151] If G=H=J, that is, the proportions of type I evaluation, type II evaluation, and type III evaluation are equal, indicating that the target publisher does not have a fixed standard when making evaluation grades, then the target publisher's evaluation preference type is recorded as four types of evaluation preferences. This classification mechanism can standardize the evaluation preferences of different publishers, so that the evaluation grades obtained by each executor are unified and standardized, which is more valuable for reference. For publishers, they can get service experience from executors that meet their preferences, which makes them more confident in subsequent use, thereby increasing the stickiness of the platform.

[0152] The task execution experience of the first type of executor is obtained, and the execution level of the first type of executor is determined in combination with the evaluation content of the second type of task published by the target publisher, specifically:

[0153] Get the evaluation level of the target publisher on the first executor, the second executor, the third executor, ... the Cth executor, which are recorded as C1, C2, C3, ... C respectively. C ;

[0154] Get C1, C2, C3, ... C C The average value is recorded as L, L = (C1 + C2 + C3 + ... + C C )÷C;

[0155] S18. Randomly select a first-class executor and obtain the number of second-class tasks executed by this type of executor in the past thirty days, and record the number as K;

[0156] S19, obtain K evaluation levels made by the publisher to this type of executor after the executor of this type performs K second-type tasks, and record them as K1, K2, K3...K K ;

[0157] S20, calculate the average value of K evaluation levels, the average value is recorded as V, V = (K1 + K2 + K3 + ... + K K )÷K;

[0158] S21. If the difference between V and L is VL>Z2, it means that the execution level of this executor is higher than the task execution standard of the target publisher, and this type of executor is recorded as a high-level executor. If the difference between V and L is LV>Z2, it means that the execution level of this executor is lower than the task execution standard of the target publisher, and this type of executor is recorded as a low-level executor. If the absolute value of the difference between V and L is |VL|≤Z2, it means that the execution level of this executor is the same as the task execution standard of the target publisher, and this type of executor is recorded as a medium executor.

[0159] Perform S18-S21 operations on all executors of the first category.

[0160] The method of pushing a target type of task to a type of executor according to the evaluation preference type of the target publisher and the evaluation type of a type of executor is specifically as follows:

[0161] If the evaluation preference type of the target publisher is a type of evaluation preference, search for medium performers in the execution area;

[0162] If there are intermediate executors in the execution area, the task push strategy is executed on the intermediate executors, specifically:

[0163] Get the distance between the location of each executor and the task location of the target task;

[0164] According to the distance between the location of each executor and the task location of the target type 1 task, the target type 1 tasks are pushed to the executors in order from near to far;

[0165] If there is no middle executor in the execution area, then search for high-level executors in the execution area. If there is a high-level executor in the execution area, then push the task strategy to the high-level executor.

[0166] If neither the high-level executor nor the medium-level executor exists in the execution area, the target type 1 task will be put into the waiting queue to wait until a medium-level executor or a high-level executor exists in the execution area;

[0167] If the evaluation preference type of the target publisher is the second-class evaluation preference, search for high-level performers in the execution area;

[0168] If there is a high-level executor in the execution area, the task push strategy is executed on the high-level executor;

[0169] If there is no high-level executor in the execution area, the target type 1 task will be placed in the waiting queue to wait until there is a high-level executor in the execution area.

[0170] The method of pushing a target task of a type to a type of executors according to the evaluation preference type of the target publisher and the evaluation type of a type of executors further includes:

[0171] If the target publisher’s evaluation preference type is recorded as three types of evaluation preference, low-level performers are retrieved in the execution area;

[0172] If there are low-level executors in the execution area, the task push strategy is executed on the low-level executors;

[0173] If there is no low-level executor in the execution area, then search for a medium-level executor in the execution area. If there is a medium-level executor in the execution area, then execute the task push strategy to the medium-level executor.

[0174] If neither the low-level executor nor the middle-level executor exists in the execution area, the high-level executor is searched in the execution area. If the high-level executor exists in the execution area, the task push strategy is executed to the high-level executor.

[0175] If there are no low-level executors, medium-level executors, or high-level executors in the execution area, the target type 1 task will be put into the waiting queue to wait until there are low-level executors and low-level executors, medium-level executors, or high-level executors in the execution area;

[0176] If the evaluation preference type of the target publisher is the four-category evaluation preference, search for high-level performers in the execution area;

[0177] If there is a high-level executor in the execution area, the task push strategy is executed on the high-level executor;

[0178] If there is no high-level executor in the execution area, then search for medium-level executors in the execution area. If there is a medium-level executor in the execution area, then execute the task push strategy to the medium-level executor.

[0179] If neither the high-level executor nor the middle-level executor exists in the execution area, the low-level executor is searched in the execution area. If the low-level executor exists in the execution area, the task push strategy is executed to the high-level executor.

[0180] If there are no low-level executors, medium-level executors, or high-level executors in the execution area, the target type 1 task will be placed in the waiting queue until there are low-level executors and low-level executors, medium-level executors, or high-level executors in the execution area. Such a grading system can not only help the platform make more accurate decisions when distributing tasks, but also motivate executors of different execution levels to improve their own levels by completing high-quality tasks; it can also ensure that task resources are more balanced and that executors of different execution levels have the same opportunities when accepting tasks;

[0181] For executors with the same execution level in the execution area, ensure that tasks are distributed to executors near the task location first; this strategy significantly improves the response speed and execution efficiency of task distribution, and is especially suitable for geographically sensitive tasks, such as on-site data collection and regional services. When a task cannot find a qualified executor in a short period of time, the platform will add the task to the waiting queue so that it can be distributed after a qualified executor appears, ensuring that high-quality tasks can be distributed to suitable executors. This mechanism improves the platform's distribution patience, so that important or high-quality tasks are more guaranteed to be completed, and also helps to improve the overall task completion rate of the platform and enhance user trust.

[0182] Embodiment 2, a system for implementing the distribution process of a mobile crowdsourcing task, is characterized by comprising:

[0183] Task content extraction module: obtains the task content of mobile crowdsourcing tasks and extracts keywords describing the task content of mobile crowdsourcing tasks;

[0184] Matching module: determines the executors who are qualified to perform the mobile crowdsourcing tasks in the execution area based on keywords;

[0185] Data processing module: determines the publisher's evaluation preference type and the executor's execution level based on the task content and evaluation content of the second-category task published by the publisher;

[0186] Task push module: pushes mobile crowdsourcing tasks to executors based on the publisher's evaluation preference type and the executor's execution level.

[0187] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0188] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the technical principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A method for distributing mobile crowdsourcing tasks, characterized in that: include, According to the publisher's evaluation preference type and the executor's execution level, the mobile crowdsourcing tasks published by the publisher are pushed to the executor, specifically: The mobile crowdsourcing tasks that have not been received and executed are recorded as the first-class tasks, and the mobile crowdsourcing tasks that have been received and executed are recorded as the second-class tasks; Randomly select a class-one task and record it as the target class-one task; The publisher of the target type 1 task is recorded as the target publisher; Get the task content of the target type task; The task content includes the task location and task content description; Establish a keyword set, recorded as set A; Extract the keywords of the task content description of the target type one task and list the keywords in set A; Determine the executors who are qualified to perform the target type of tasks based on the keywords in set A; Retrieve the executors who are qualified to execute the target type one task in the execution area and record them as type one executors; Obtain the task content and evaluation content of the second-category task published by the target publisher; Determine the evaluation preference type of the target publisher according to the task content and evaluation content of the second type of tasks published by the target publisher; Obtain the task execution experience of the first-class executors, and determine the execution level of the first-class executors in combination with the evaluation content of the second-class tasks published by the target publisher; According to the evaluation preference type of the target publisher and the execution level of the first type of executors, a target type of tasks is pushed to the first type of executors.

2. A method for distributing mobile crowdsourcing tasks according to claim 1, characterized in that: The executors who have the qualification to execute the target type one task in the search execution area are recorded as type one executors, specifically: The area selected by the target publisher when publishing the target type task is recorded as the execution area; Obtain the task execution experience of all executors in the execution area; The task execution experience includes the task content of the task executed and the evaluation level made by the publisher after the task is executed; The evaluation level is specifically 0, 100, and a whole number within 100, wherein the larger the number, the higher the evaluation level; Set the keyword proportion threshold Z1; S1. Randomly select an executor in the execution area; S2. Obtain the task content description of all task execution experiences of this executor; S3, extracting keywords describing the task content in all the task execution experiences of this executor; S4, establish a keyword set, record it as set B, and put the extracted keywords into set B; S5. Randomly select a keyword in set A. If the selected keyword in set A exists in set B, the selected keyword in set A is recorded as a matching keyword; S6. Obtain the number of keywords and matching keywords in set A respectively, record them as A1 and A2 respectively, and calculate the ratio of A2 to A1, record them as A2÷A1; S7. If A2÷A1≥Z1, then this executor is qualified to receive and execute the target type 1 task; S8. If A2÷A1<Z1, then this executor is not qualified to receive and execute the target type 1 task; Perform S1-S8 operation steps on all executors in the execution area to determine all executors in the execution area who are qualified to execute the target type one task, and record all executors in the execution area who are qualified to execute the target type one task as type one executors.

3. The method for distributing mobile crowdsourcing tasks according to claim 1, characterized in that: The evaluation preference type of the target publisher is determined according to the task content and evaluation content of the second type of tasks published by the target publisher, specifically: Get the number of executors who have executed the second-class tasks issued by the target publisher in the past thirty days, and record the number as C; The C executors who execute the second-category tasks issued by the target publisher within nearly thirty days are respectively recorded as the first executor, the second executor, the third executor, and the Cth executor; Obtain the task execution experience of the first executor, the second executor, the third executor, ... the Cth executor in sequence; S9. Select any one executor from the first executor, the second executor, the third executor, ... the Cth executor: S10, obtaining the number of the second-category tasks executed by the executor in the past thirty days, the number being recorded as D; S11, respectively obtain the evaluation level of the executor after the executor performs D second-class tasks, and the evaluation levels are recorded as D1, D2, D3...D D ; S12, obtaining the evaluation level made by the target publisher after the executor performs the second type of task of the target publisher, where the evaluation level is recorded as d; S13, obtain D1, D2, D3...D respectively D The absolute value of the difference from d; S14, setting a difference threshold Z2, and setting an evaluation level proportion threshold Z3; S15, if D1, D2, D3...D D If the ratio of the number of absolute values ​​of the difference between the values ​​of Z2 and d is greater than or equal to Z3, the evaluation type of the target publisher on this executor is recorded as a first-class evaluation; S16, if D1, D2, D3...D D Among the absolute values ​​of the differences from d, if the ratio of the number of absolute values ​​less than or equal to Z2 to D is less than Z3, then D1, D2, D3, ..., D are obtained respectively. D The number of evaluation levels with values ​​greater than d and less than d are recorded as E and F respectively; S17. If E≥F, the target publisher's evaluation type for the executor is recorded as a second-class evaluation; if E<F, the target publisher's evaluation type for the executor is recorded as a third-class evaluation; The operations S9 to S17 are performed on the first executor, the second executor, the third executor, ... the Cth executor in sequence to obtain the evaluation types of the target publisher on the first executor, the second executor, the third executor, ... the Cth executor.

4. The method for distributing mobile crowdsourcing tasks according to claim 1, characterized in that: The step of determining the evaluation preference type of the target publisher according to the task content and evaluation content of the second type of task published by the target publisher also includes: Obtain the number of first-category evaluations, second-category evaluations, and third-category evaluations of the C evaluation types of the target publisher on the first executor, the second executor, the third executor, ..., the Cth executor, respectively, and record the numbers as G, H, and J respectively; If G ≥ H and G > J, or G > H and G ≥ J, the target publisher's evaluation preference type is recorded as a type of evaluation preference; If H>G and H≥J, the target publisher’s evaluation preference type is recorded as the second-class evaluation preference; If J>G and J>H, the target publisher’s evaluation preference type is recorded as three types of evaluation preference; If G=H=J, the evaluation preference types of the target publisher are recorded as four types of evaluation preferences.

5. The method for distributing mobile crowdsourcing tasks according to claim 1, characterized in that: The task execution experience of the first type of executor is obtained, and the execution level of the first type of executor is determined in combination with the evaluation content of the second type of task published by the target publisher, specifically: Get the evaluation level of the target publisher on the first executor, the second executor, the third executor, ... the Cth executor, which are recorded as C1, C2, C3, ... C respectively. C ; Get C1, C2, C3, ... C C The average value is recorded as L, L=(C1+C2+C3+……+C C )÷C; S18. Randomly select a first-class executor and obtain the number of second-class tasks executed by this type of executor in the past thirty days, and record the number as K; S19, obtain K evaluation levels made by the publisher to this type of executor after the executor of this type performs K second-type tasks, and record them as K1, K2, K3...K K ; S20. Calculate the average value of K evaluation levels, and record the average value as V, V=(K1+K2+K3+……+K K )÷K; S21. If the difference between V and L VL>Z2, where Z2 is the difference threshold, then this type of executor is recorded as a high-level executor; if the difference between V and L LV>Z2, then this type of executor is recorded as a low-level executor; if the absolute value of the difference between V and L |VL|≤Z2, then this type of executor is recorded as a medium-level executor; Perform S18-S21 operations on all executors of the first category.

6. A method for distributing mobile crowdsourcing tasks according to claim 1, characterized in that: The method of pushing a target type of task to a type of executor according to the evaluation preference type of the target publisher and the evaluation type of a type of executor is specifically as follows: If the evaluation preference type of the target publisher is a type of evaluation preference, search for medium performers in the execution area; If there are intermediate executors in the execution area, the task push strategy is executed on the intermediate executors, specifically: Get the distance between the location of each executor and the task location of the target task; According to the distance between the location of each executor and the task location of the target type 1 task, the target type 1 tasks are pushed to the executors in order from near to far; If there is no middle executor in the execution area, then search for high-level executors in the execution area. If there is a high-level executor in the execution area, then push the task strategy to the high-level executor. If neither the high-level executor nor the medium-level executor exists in the execution area, the target type 1 task will be put into the waiting queue to wait until a medium-level executor or a high-level executor exists in the execution area; If the evaluation preference type of the target publisher is the second-class evaluation preference, search for high-level performers in the execution area; If there is a high-level executor in the execution area, the task push strategy is executed on the high-level executor; If there is no high-level executor in the execution area, the target type 1 task will be placed in the waiting queue to wait until there is a high-level executor in the execution area.

7. A method for distributing mobile crowdsourcing tasks according to claim 1, characterized in that: The method of pushing a target task of a type to a type of executors according to the evaluation preference type of the target publisher and the evaluation type of a type of executors further includes: If the target publisher’s evaluation preference type is recorded as three types of evaluation preference, low-level performers are retrieved in the execution area; If there are low-level executors in the execution area, the task push strategy is executed on the low-level executors; If there is no low-level executor in the execution area, then search for a medium-level executor in the execution area. If there is a medium-level executor in the execution area, then execute the task push strategy to the medium-level executor. If neither the low-level executor nor the middle-level executor exists in the execution area, the high-level executor is searched in the execution area. If the high-level executor exists in the execution area, the task push strategy is executed on the high-level executor. If there are no low-level executors, medium-level executors, or high-level executors in the execution area, the target type 1 task will be put into the waiting queue to wait until there are low-level executors and low-level executors, medium-level executors, or high-level executors in the execution area; If the evaluation preference type of the target publisher is the four-category evaluation preference, search for high-level performers in the execution area; If there is a high-level executor in the execution area, the task push strategy is executed on the high-level executor; If there is no high-level executor in the execution area, then search for medium-level executors in the execution area. If there is a medium-level executor in the execution area, then execute the task push strategy to the medium-level executor. If neither the high-level executor nor the middle-level executor exists in the execution area, the low-level executor is searched in the execution area. If the low-level executor exists in the execution area, the task push strategy is executed to the high-level executor. If there are no low-level executors, medium-level executors, or high-level executors in the execution area, the target type 1 task will be placed in a waiting queue and wait until there are low-level executors and low-level executors, medium-level executors, or high-level executors in the execution area.

8. A system for implementing the method for distributing a mobile crowdsourcing task as claimed in claim 1, characterized in that: include, Task content extraction module: obtains the task content of mobile crowdsourcing tasks and extracts keywords describing the task content of mobile crowdsourcing tasks; Matching module: determines the executors who are qualified to perform the mobile crowdsourcing tasks in the execution area based on keywords; Data processing module: determines the publisher's evaluation preference type and the executor's execution level based on the task content and evaluation content of the second-category task published by the publisher; Task push module: pushes mobile crowdsourcing tasks to executors based on the publisher's evaluation preference type and the executor's execution level.

Images