CN110443502A - Crowdsourcing task recommendation method and system based on worker's capability comparison - Google Patents

Crowdsourcing task recommendation method and system based on worker's capability comparison Download PDF

Info

Publication number
CN110443502A
CN110443502A CN201910723106.7A CN201910723106A CN110443502A CN 110443502 A CN110443502 A CN 110443502A CN 201910723106 A CN201910723106 A CN 201910723106A CN 110443502 A CN110443502 A CN 110443502A
Authority
CN
China
Prior art keywords
worker
task
data
feature vector
crowdsourcing
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN201910723106.7A
Other languages
Chinese (zh)
Inventor
彭张林
陈丹红
张强
陆效农
陈萍
王安宁
廖婧萍
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Hefei University of Technology
Hefei Polytechnic University
Original Assignee
Hefei Polytechnic University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Hefei Polytechnic University filed Critical Hefei Polytechnic University
Priority to CN201910723106.7A priority Critical patent/CN110443502A/en
Publication of CN110443502A publication Critical patent/CN110443502A/en
Pending legal-status Critical Current

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G06Q10/063—Operations research, analysis or management
    • G06Q10/0631—Resource planning, allocation, distributing or scheduling for enterprises or organisations
    • G06Q10/06311—Scheduling, planning or task assignment for a person or group
    • G06Q10/063112—Skill-based matching of a person or a group to a task
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/10—Office automation; Time management
    • G06Q10/103—Workflow collaboration or project management
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q30/00—Commerce
    • G06Q30/06—Buying, selling or leasing transactions
    • G06Q30/08—Auctions

Landscapes

  • Business, Economics & Management (AREA)
  • Human Resources & Organizations (AREA)
  • Engineering & Computer Science (AREA)
  • Strategic Management (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Economics (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Marketing (AREA)
  • General Physics & Mathematics (AREA)
  • General Business, Economics & Management (AREA)
  • Tourism & Hospitality (AREA)
  • Quality & Reliability (AREA)
  • Operations Research (AREA)
  • Educational Administration (AREA)
  • Development Economics (AREA)
  • Accounting & Taxation (AREA)
  • Finance (AREA)
  • Data Mining & Analysis (AREA)
  • Game Theory and Decision Science (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

The present invention provides a kind of crowdsourcing task recommendation method and system based on worker's capability comparison, is related to data processing field.The present invention obtains the feature vector table of worker's ability of fusion qualified quality model by task data on crowdsourcing platform and worker data, worker similar with target worker's ability is found out by feature vector table, the task bid record for being then based on similar worker generates preliminary task recommendation list, finally based on the task bid record for generating preliminary task recommendation list and target worker, final task recommendation list is generated, and gives the crowdsourcing task recommendation in the final task recommendation list to target worker.The present invention carries out task recommendation from the ability angle of worker, according to the task bid record of similar worker, effectively reduces time cost and search cost that worker finds task on crowdsourcing platform.

Description

Crowdsourcing task recommendation method and system based on worker's capability comparison
Technical field
The present invention relates to technical field of data processing, and in particular to a kind of crowdsourcing task recommendation based on worker's capability comparison Method and system.
Background technique
Crowdsourcing refers to the task that a company or mechanism are executed the past by employee, in freely voluntary form outside It wraps to the way of unspecific public network.Under crowdsourcing model, the problem of oneself can be passed through network crowdsourcing platform by enterprise Publication, and certain fund is provided and solves the problems, such as enterprise to recruit the talent.Solutionist can seek on crowdsourcing platform simultaneously Being suitble to oneself of the task is looked for, submits and just has an opportunity to obtain certain remuneration after solving achievement.This mode reduce enterprise at This, while meeting supply and demand double the needs of sending out, there is good development prospect.
The basic procedure of crowdsourcing task are as follows: task is published to crowdsourcing platform by task publisher, and provides corresponding remuneration； The worker of crowdsourcing platform may browse through task, if their tasks interested that can register；Task publisher is from numerous Suitable worker is selected to develop the task in registration worker；Worker successfully completes task and obtains corresponding remuneration.From process It is found that worker finds how the properly task of oneself and task publisher select suitable worker to open from numerous registration workers The task is sent out, is all the recommender system problem in need of consideration in crowdsourcing model.
However, finding this problem of the suitable task of oneself to worker, existing crowdsourcing model recommended method is only from crowdsourcing Task characteristic angle considers worker's ability, is unfavorable for worker and finds suitable task on crowdsourcing platform, causes worker in crowdsourcing The waste of task excessive time and cost are found on platform.
Summary of the invention
(1) the technical issues of solving
In view of the deficiencies of the prior art, the crowdsourcing task recommendation method based on worker's capability comparison that the present invention provides a kind of And system, solving existing crowdsourcing task recommendation method will lead to worker when the waste for finding task on crowdsourcing platform is excessive Between and the technical issues of cost.
(2) technical solution
In order to achieve the above object, the present invention is achieved by the following technical programs:
The present invention provides a kind of crowdsourcing task recommendation method based on worker's capability comparison, and the method is held by computer Row, which comprises
Task data and worker data on S1, acquisition crowdsourcing platform, are based on the task data and the worker data The task bid record of worker is obtained, the task bid record based on worker screens worker data；
Worker data after S2, task based access control data and screening obtains the feature of worker's ability of fusion qualified quality model Vector table；
S3, similar worker similar to target worker's ability is obtained based on described eigenvector table；
S4, the task bid record based on similar worker generate preliminary task recommendation list；
S5, the task bid record based on generation preliminary the task recommendation list and target worker, generate final task Recommendation list, and give the crowdsourcing task recommendation in the final task recommendation list to target worker.
Preferably, in S1, the screening technique of the worker data are as follows:
Removing task bid record is less than or equal to the worker data of the worker of M item, retains task bid record and is greater than M item Worker worker data.
Preferably, the step S2 includes:
Feature vector of the worker in technical ability is obtained based on the worker data after screening；
Worker data after task based access control data and screening obtains worker in intellectual feature vector；
Worker data after task based access control data and screening obtains feature vector of the worker in self-concept；
Feature vector of the worker in speciality is obtained based on the worker data after screening；
Feature vector of the worker in motivation is obtained based on the worker data after screening；
In conjunction with feature vector of the worker in technical ability, worker in intellectual feature vector, worker in self-concept The feature vector of feature vector and worker in motivation of feature vector, worker in speciality obtains described eigenvector table.
Preferably, the step S3 includes:
Data in described eigenvector table are clustered, in cluster process are as follows: to feature vector in feature vector table Dimensionality reduction is carried out, K-means cluster is carried out to the feature vector after dimensionality reduction, obtains cluster result；
Similar worker similar to target worker's ability is found according to cluster result.
Preferably, the step S4 includes:
The task bid record of similar worker is obtained from the task bid record of worker；
Preliminary task recommendation list is generated based on the task bid record, recommends the formula of principle as follows:
Wherein:
TirIt indicates to target worker uiRecommendation set；
Indicate target worker uiAll similar workers whole bid set of tasks；
TaIndicate worker uiBid set of tasks.
Preferably, the step S5 includes:
The task bid record of target worker is obtained from the task bid record of worker；
What the task bid record prediction target worker based on target worker selected task in preliminary task recommendation list can Energy property, taking possibility is the task of preceding N, and N value can be preset, and generates final task recommendation list, top n crowdsourcing task is pushed away It recommends and gives target worker, the calculation formula of possibility is as follows:
Wherein:
U refers to worker u；
K is the totality that k worker is formed；
Cw,jIndicate similar worker uwTo the bid intention of crowdsourcing task j, that is, work as uwWhen having task bid record to j, Cw,j=1, otherwise, Cw,j=0；
Sim (i, w)=1.
Preferably, the worker data include: worker's title, grade, positive rating, speed, quality, attitude, collection number, Transaction total amount, acceptance of the bid total degree, retail shop improve rate and are good at technical ability；The task data includes: task names, task report Reward, task category, task participate in number, highest bidder and bidder.
Preferably, the acquisition methods of feature vector of the worker in technical ability specifically include:
Feature vector of the worker in technical ability is obtained based on the technical ability of being good in worker data, use 0 indicates that worker does not have A certain to be good at technical ability, use 1 indicates that worker has and a certain is good at technical ability, worker uiFeature vector in technical ability may be expressed as:
Ki=(k1,i,k2,i,k3,i,…,ky,i)；
The worker specifically includes in the acquisition methods of intellectual feature vector:
Task category in task bid record based on worker and task data obtains the task class that worker participates in task Not；
Worker is obtained in intellectual feature vector based on the task category that worker participates in task；
Indicate that worker does not have a certain knowledge with 0, use 1 indicates that worker has a certain knowledge, worker uiIn intellectual spy Sign vector may be expressed as:
Si=(s1,i,s2,i,s3,i,…,sx,i)；
The acquisition methods of feature vector of the worker in self-concept specifically include:
Worker data obtains task averagely participation amount after task based access control data and screening；
Retail shop in task based access control averagely participation amount and worker data improves rate and obtains feature of the worker in self-concept Vector, worker uiFeature vector in self-concept may be expressed as:
Ai=(pi,ri)
Wherein:
piFor worker uiTask averagely participation amount；
riFor worker uiRetail shop improve rate；
The acquisition methods of feature vector of the worker in speciality specifically include:
Worker is obtained in speciality based on use grade, positive rating, speed, quality, attitude and the collection number in worker data On feature vector, worker uiFeature vector in speciality may be expressed as:
Hi=(gi, ci, di, qi, ai, ni)
Wherein:
giFor worker uiGrade；
ciFor worker uiPositive rating；
diFor worker uiSpeed；
qiFor worker uiQuality；
aiFor worker uiAttitude；
niFor worker uiCollection number；
The acquisition methods of feature vector of the worker in motivation specifically include:
Feature vector of the worker in motivation, worker are obtained based on the acceptance of the bid total degree in worker data and total amount of trading uiFeature vector in motivation may be expressed as:
Mi=(ti,bi)
Wherein:
tiFor worker uiAcceptance of the bid total degree；
biFor worker uiTransaction total amount.
The crowdsourcing task recommendation system based on worker's capability comparison that the present invention also provides a kind of, which is characterized in that the system System includes computer, and the computer includes:
At least one storage unit；
At least one processing unit；
Wherein, at least one instruction is stored at least one described storage unit, at least one instruction is by described At least one processing unit is loaded and is executed to perform the steps of
Task data and worker data on S1, acquisition crowdsourcing platform, are based on the task data and the worker data The task bid record of worker is obtained, the task bid record based on worker screens worker data；
Worker data after S2, task based access control data and screening obtains the spy of worker's ability of fusion qualified quality model Levy vector table；
S3, similar worker similar to target worker's ability is obtained based on described eigenvector table；
S4, the task bid record based on similar worker generate preliminary task recommendation list；
S5, the task bid record based on generation preliminary the task recommendation list and target worker, generate final task Recommendation list, and give the crowdsourcing task recommendation in the final task recommendation list to target worker.
(3) beneficial effect
The present invention provides a kind of crowdsourcing task recommendation method and system based on worker's capability comparison.With prior art phase Than, have it is following the utility model has the advantages that
The present invention obtains worker's energy of fusion qualified quality model by task data on crowdsourcing platform and worker data The feature vector table of power finds out worker similar with target worker's ability by feature vector table, is then based on similar worker's Task bid record generates preliminary task recommendation list, finally generates preliminary task recommendation list and target worker based on described Task bid record generates final task recommendation list, and the crowdsourcing task recommendation in the final task recommendation list is given Target worker.The present invention carries out task recommendation from the ability angle of worker, according to similar worker's task bid record, effectively Reduction worker the time cost and search cost of task are found on crowdsourcing platform.
Detailed description of the invention
In order to more clearly explain the embodiment of the invention or the technical proposal in the existing technology, to embodiment or will show below There is attached drawing needed in technical description to be briefly described, it should be apparent that, the accompanying drawings in the following description is only this Some embodiments of invention for those of ordinary skill in the art without creative efforts, can be with It obtains other drawings based on these drawings.
Fig. 1 is a kind of block diagram of the crowdsourcing task recommendation method based on worker's capability comparison of the embodiment of the present invention.
Specific embodiment
In order to make the object, technical scheme and advantages of the embodiment of the invention clearer, to the technology in the embodiment of the present invention Scheme is clearly and completely described, it is clear that and described embodiments are some of the embodiments of the present invention, rather than whole Embodiment.Based on the embodiments of the present invention, those of ordinary skill in the art are obtained without creative efforts The every other embodiment obtained, shall fall within the protection scope of the present invention.
The embodiment of the present application is solved by providing a kind of crowdsourcing task recommendation method and system based on worker's capability comparison Existing crowdsourcing task recommendation method will lead to waste excessive time and cost that worker finds on crowdsourcing platform task Problem is realized and carries out task recommendation according to similar worker's task bid record, appoints to reduce worker and find on crowdsourcing platform The time cost and search cost of business.
Technical solution in the embodiment of the present application is in order to solve the above technical problems, general thought is as follows:
The embodiment of the present invention is found out and target worker by merging the feature vector table of worker's ability of qualified quality model The similar similar worker of ability, then the task bid record of similar worker is handled, it ultimately generates final task and recommends column Table gives crowdsourcing task recommendation to target worker.The embodiment of the present invention is from the ability angle of worker, according to similar worker's task Bid record carries out task recommendation, to reduce time cost and search cost that worker finds task on crowdsourcing platform.
In order to better understand the above technical scheme, in conjunction with appended figures and specific embodiments to upper Technical solution is stated to be described in detail.
The crowdsourcing task recommendation method based on worker's capability comparison that the embodiment of the invention provides a kind of, this method is by calculating Machine executes, as shown in Figure 1, the method comprising the steps of S1~S5:
Task data and worker data on S1, acquisition crowdsourcing platform, are based on the task data and the worker data The task bid record of worker is obtained, the task bid record based on worker screens worker data；
Worker data after S2, task based access control data and screening obtains the feature of worker's ability of fusion qualified quality model Vector table；
S3, similar worker similar to target worker's ability is obtained based on described eigenvector table；
S4, the task bid record based on similar worker generate preliminary task recommendation list；
S5, the task bid record based on generation preliminary the task recommendation list and target worker, generate final task Recommendation list, and give the crowdsourcing task recommendation in the final task recommendation list to target worker.
The embodiment of the present invention from the feature vector table of worker's ability of fusion qualified quality model by finding out and target The similar similar worker of worker's ability, then the task bid record of similar worker is handled, it ultimately generates final task and pushes away List is recommended, gives crowdsourcing task recommendation to target worker.The embodiment of the present invention is from the ability angle of worker, according to similar worker Task bid record carry out task recommendation, thus reduce worker is found on crowdsourcing platform task time cost and search at This.
Each step is described in detail below.
In step sl, the task data and worker data on crowdsourcing platform are obtained, based on the task data and described Worker data obtains the task bid record of worker, and the task bid record based on worker screens worker data.Specifically Implementation process it is as follows:
S101, task data and worker data on crowdsourcing platform are obtained by web crawlers technology.Wherein, worker data packet Include: worker's title, grade, positive rating, speed, quality, attitude, collection number, transaction total amount, acceptance of the bid total degree, retail shop are complete Kind rate is good at technical ability；The task data includes: task names, task remuneration, task category, task participation number, acceptance of the bid People, bidder.
S102, worker data is screened, the method for screening is as follows:
Task bid record table is constructed according to task data and worker data, such as table 1, it should be noted that table 1 is to appoint The local content of business bid record table.
Table 1
Worker's number Mission number
1 290
1 32
1 99
1 99
1 224
1 30
1 312
1 81
1 336
1 348
1 24
1 15
1 79
1 187
In table 1, table first item is worker's number, and Section 2 corresponds to the number for the task that worker submitted a tender, from task Analysis obtains the task bid record of worker in bid record table；Because bid record is very few, it is unfavorable for obtaining worker's feature, shadow Subsequent task recommendation is rung, so removing the worker data that task bid record is less than or equal to the worker of M item, retains task and throws Label record is greater than the worker data of the worker of M item.In embodiments of the present invention, M=5.
In step s 2, task based access control data obtain worker's energy of fusion qualified quality model with the worker data after screening The feature vector table of power.Qualified quality model be individual for a certain work of completion, reach that a certain performance objective should have be The combination for arranging different quality elements is divided into several aspects such as intrinsic motivation, knowledge expertise, self-image and social role's feature. Specific implementation process is as follows:
S201, acquisition five feature vectors relevant to worker's ability.It is specific as follows:
S2011, feature vector of the worker in technical ability, specific process are obtained based on the worker data after screening are as follows: Feature vector of the worker in technical ability is obtained based on the technical ability of being good in worker data, use 0 indicates that worker does not have a certain be good at Technical ability, use 1 indicate that worker has and a certain are good at technical ability, worker uiFeature vector in technical ability may be expressed as:
Ki=(k1,i,k2,i,k3,i,…,ky,i)
It should be noted that in embodiments of the present invention, the technical ability that worker is good at is divided into 6 classes (i.e. y=6), such as software Develop the technical ability such as class, interface class, data analysis classes.
Worker data after S2012, task based access control data and screening obtains worker in intellectual feature vector, specifically Process are as follows:
Task category in task bid record based on worker and task data obtains the task class that worker participates in task Not；
Worker is obtained in intellectual feature vector based on the task category that worker participates in task；
Indicate that worker does not have a certain knowledge with 0, use 1 indicates that worker has a certain knowledge, and worker ui is in intellectual spy Sign vector may be expressed as:
Si=(s1,i,s2,i,s3,i,…,sx,i)；
It should be noted that in embodiments of the present invention, the knowledge that worker grasps or has is divided into 9 classes (i.e. x=9).
S2013, feature vector of the worker in self-concept is obtained based on the worker data after screening.In competency mould In type, self-concept is the description of the attitude about a people, value, self-image.Whether the embodiment of the present invention is found pleasure in worker In receiving that there is competitive and challenge task, and to two dimensions of cognition of self-image come measure worker self is general It reads.Wherein, measure whether worker is partial to competitive, challenge task, from work with worker's task averagely participation amount The retail shop of people individual improves rate to measure cognition of the worker to self-image.Detailed process is as follows:
Worker data obtains (the i.e. participation of the task of worker's participation of task averagely participation amount after task based access control data and screening Number is averaged)；
Retail shop in task based access control averagely participation amount and worker data improves rate and obtains feature of the worker in self-concept Vector, worker uiFeature vector in self-concept may be expressed as:
Ai=(pi,ri)
Wherein:
piFor worker uiTask averagely participation amount；
riFor worker uiRetail shop improve rate；
S2014, feature vector of the worker in speciality is obtained based on the worker data after screening.Speciality be a people for Ambient enviroment and the lasting reaction of information.In crowdsourcing platform, this reaction is interpreted as worker, itself and task are wanted Seek embodied Some features.Such as grade, positive rating, speed, quality, attitude and collection number, these indexs are all durations The embodiment of reaction.Detailed process is as follows:
Worker is obtained in speciality based on use grade, positive rating, speed, quality, attitude and the collection number in worker data On feature vector, worker uiFeature vector in speciality may be expressed as:
Hi=(gi, ci, di, qi, ai, ni)
Wherein:
giFor worker uiGrade；
ciFor worker uiPositive rating；
diFor worker uiSpeed；
qiFor worker uiQuality；
aiFor worker uiAttitude；
niFor worker uiCollection number.
S2015, feature vector of the worker in motivation is obtained based on the worker data after screening.Motivation be determine individual and An important factor for organizational behavior, be excitation worker participate in crowdsourcing task an important factor for, motivation can be divided into again external sexual motivation with Internal sexual motivation.The bonus total amount that the embodiment of the present invention mainly uses worker's transaction to obtain obtains this to measure worker to money It is dynamic to professional honor or this internal property of sense of ownership to measure worker with the total degree of worker's acceptance of the bid task for one outside sexual motivation Machine.Detailed process is as follows:
Feature vector of the worker in motivation, worker are obtained based on the acceptance of the bid total degree in worker data and total amount of trading uiFeature vector in motivation may be expressed as:
Mi=(ti,bi)
Wherein:
tiFor worker uiAcceptance of the bid total degree；
biFor worker uiTransaction total amount.
S202, in conjunction with feature vector of the worker in technical ability, worker in intellectual feature vector, worker in self-concept On the feature vector of feature vector and worker in motivation in speciality of feature vector, worker, obtain described eigenvector Table.Feature vector table is shown in Table 2, it should be noted that table 2 is characterized the local content of vector table.
Table 2
In step s3, similar worker similar to target worker's ability is obtained based on described eigenvector table.It is specific real It is as follows to apply process:
S301, feature vector table is clustered, in cluster process are as follows:
Dimensionality reduction is carried out to feature vector in feature vector table, is down to 2 dimensions from 25 dimensions.Feature vector after dimensionality reduction is carried out K-means cluster, obtains cluster result.
S303, similar worker similar to target worker's ability is found according to cluster result.
In step s 4, the task bid record based on similar worker generates preliminary task recommendation list.It was embodied Journey is as follows:
S401, it is numbered according to the worker of similar worker, the task of similar worker is obtained from the task bid record of worker Bid record；
S402, preliminary task recommendation list is generated based on the task bid record, recommends the formula of principle as follows:
Wherein:
TI coverIt indicates to target worker uiRecommendation set；
TciIndicate target worker uiAll similar workers whole bid set of tasks；
TaIndicate worker uiBid set of tasks.
It should be noted that in embodiments of the present invention, to participation amount ranking in certain a kind of worker, first five worker is carried out Task recommendation generates preliminary task recommendation list.
In step s 5, it based on the task bid record for generating preliminary task recommendation list and target worker, generates Final task recommendation list, and give the crowdsourcing task recommendation in the final task recommendation list to target worker.Specific implementation Process is as follows:
S501, it is numbered according to the worker of target worker, the task of target worker is obtained from the task bid record of worker Bid record；
S502, the task bid record prediction target worker based on target worker select task in preliminary task recommendation list A possibility that, taking possibility is Top-N task, generates final task recommendation list, gives top n crowdsourcing task recommendation to target work The calculation formula of people, possibility are as follows:
Wherein:
U refers to worker u；
K is the totality that k worker is formed；
Cw,jIndicate similar worker uwTo the bid intention of crowdsourcing task j, that is, work as uwWhen having task bid record to j, Cw,j=1, otherwise, Cw,j=0；
Sim (i, w)=1.
The embodiment of the present invention selects possibility to generate final task recommendation list for the task of preceding 5 (i.e. N=5), specifically As shown in table 3.
Table 3
The embodiment of the present invention also provides a kind of crowdsourcing task recommendation system based on worker's capability comparison, the system comprises Computer, the computer include:
At least one storage unit；
At least one processing unit；
Wherein, at least one instruction is stored at least one described storage unit, at least one instruction is by described At least one processing unit is loaded and is executed to perform the steps of
Task data and worker data on S1, acquisition crowdsourcing platform, are based on the task data and the worker data The task bid record of worker is obtained, the task bid record based on worker screens worker data；
Worker data after S2, task based access control data and screening obtains the feature of worker's ability of fusion qualified quality model Vector table；
S3, similar worker similar to target worker's ability is obtained based on described eigenvector table；
S4, the task bid record based on similar worker generate preliminary task recommendation list；
S5, the task bid record based on generation preliminary the task recommendation list and target worker, generate final task Recommendation list, and give the crowdsourcing task recommendation in the final task recommendation list to target worker.
In conclusion compared with prior art, have it is following the utility model has the advantages that
1, the embodiment of the present invention pass through to fusion qualified quality model worker's ability feature vector table in data into Row cluster finds out similar worker similar to target worker's ability according to cluster result, then submits a tender and remember to the task of similar worker Record is handled, and is ultimately generated final task recommendation list, is given crowdsourcing task recommendation to target worker.The embodiment of the present invention is from work The ability angle of people is set out, and task recommendation is carried out according to similar worker's task bid record, to reduce worker in crowdsourcing platform The time cost and search cost of upper searching task.
2, the final task recommendation list that the embodiment of the present invention ultimately generates be based on crowdsourcing platform worker data and What task data obtained, data directly are obtained from crowdsourcing platform, not only obtains and is easy, and can ensure real result.
3, the embodiment of the present invention is clustered according to the feature vector table of worker's ability of fusion qualified quality model, favorably In accurately searching out similar worker.
It should be noted that through the above description of the embodiments, those skilled in the art can be understood that It can be realized by means of software and necessary general hardware platform to each embodiment.Based on this understanding, above-mentioned skill Substantially the part that contributes to existing technology can be embodied in the form of software products art scheme in other words, the calculating Machine software product may be stored in a computer readable storage medium, such as ROM/RAM, magnetic disk, CD, including some instructions are used So that computer equipment (can be personal computer, server or the network equipment etc.) execute each embodiment or Method described in certain parts of person's embodiment.
Herein, relational terms such as first and second and the like be used merely to by an entity or operation with it is another One entity or operation distinguish, and without necessarily requiring or implying between these entities or operation, there are any this reality Relationship or sequence.Moreover, the terms "include", "comprise" or its any other variant are intended to the packet of nonexcludability Contain, so that the process, method, article or equipment for including a series of elements not only includes those elements, but also including Other elements that are not explicitly listed, or further include for elements inherent to such a process, method, article, or device. In the absence of more restrictions, the element limited by sentence "including a ...", it is not excluded that including the element Process, method, article or equipment in there is also other identical elements.
The above embodiments are merely illustrative of the technical solutions of the present invention, rather than its limitations；Although with reference to the foregoing embodiments Invention is explained in detail, those skilled in the art should understand that: it still can be to aforementioned each implementation Technical solution documented by example is modified or equivalent replacement of some of the technical features；And these modification or Replacement, the spirit and scope for technical solution of various embodiments of the present invention that it does not separate the essence of the corresponding technical solution.

Claims (9)

1. a kind of crowdsourcing task recommendation method based on worker's capability comparison, which is characterized in that the method is executed by computer, The described method includes:
Task data and worker data on S1, acquisition crowdsourcing platform, are obtained based on the task data and the worker data The task bid record of worker, the task bid record based on worker screen worker data；
Worker data after S2, task based access control data and screening obtains the feature vector of worker's ability of fusion qualified quality model Table；
S3, similar worker similar to target worker's ability is obtained based on described eigenvector table；
S4, the task bid record based on similar worker generate preliminary task recommendation list；
S5, the task bid record based on generation preliminary the task recommendation list and target worker, generate final task and recommend List, and give the crowdsourcing task recommendation in the final task recommendation list to target worker.
2. the crowdsourcing task recommendation method based on worker's capability comparison as described in claim 1, which is characterized in that in S1, The screening technique of the worker data are as follows:
Removing task bid record is less than or equal to the worker data of the worker of M item, retains the work that task bid record is greater than M item The worker data of people.
3. the crowdsourcing task recommendation method based on worker's capability comparison as described in claim 1, which is characterized in that the step S2 includes:
Feature vector of the worker in technical ability is obtained based on the worker data after screening；
Worker data after task based access control data and screening obtains worker in intellectual feature vector；
Worker data after task based access control data and screening obtains feature vector of the worker in self-concept；
Feature vector of the worker in speciality is obtained based on the worker data after screening；
Feature vector of the worker in motivation is obtained based on the worker data after screening；
In conjunction with feature vector, worker feature in intellectual feature vector, worker on self-concept of the worker in technical ability The feature vector of feature vector and worker in motivation of vector, worker in speciality obtains described eigenvector table.
4. the crowdsourcing task recommendation method based on worker's capability comparison as claimed in claim 3, which is characterized in that the step S3 includes:
Data in described eigenvector table are clustered, in cluster process are as follows: feature vector in feature vector table is carried out Dimensionality reduction carries out K-means cluster to the feature vector after dimensionality reduction, obtains cluster result；
Similar worker similar to target worker's ability is found according to cluster result.
5. the crowdsourcing task recommendation method based on worker's capability comparison as claimed in claim 4, which is characterized in that the step S4 includes:
The task bid record of similar worker is obtained from the task bid record of worker；
Preliminary task recommendation list is generated based on the task bid record, recommends the formula of principle as follows:
Wherein:
TirIt indicates to target worker uiRecommendation set；
Indicate target worker uiAll similar workers whole bid set of tasks；
TaIndicate worker uiBid set of tasks.
6. the crowdsourcing task recommendation method based on worker's capability comparison as claimed in claim 5, which is characterized in that the step S5 includes:
The task bid record of target worker is obtained from the task bid record of worker；
A possibility that task bid record prediction target worker based on target worker selects task in preliminary task recommendation list, Taking possibility is the task of preceding N, and N value can be preset, and generates final task recommendation list, top n crowdsourcing task recommendation is given The calculation formula of target worker, possibility are as follows:
Wherein:
U refers to worker u；
K is the totality that k worker is formed；
CW, jIndicate similar worker uwTo the bid intention of crowdsourcing task j, that is, work as uwWhen having task bid record to j, CW, j= 1, otherwise, CW, j=0；
Sim (i, w)=1.
7. the crowdsourcing task recommendation method based on worker's capability comparison as claimed in claim 3, which is characterized in that the worker Data include: worker's title, grade, positive rating, speed, quality, attitude, collection number, transaction total amount, acceptance of the bid total degree, Retail shop improves rate and is good at technical ability；The task data includes: task names, task remuneration, task category, task participant Number, highest bidder and bidder.
8. the crowdsourcing task recommendation method based on worker's capability comparison as claimed in claim 7, which is characterized in that the worker The acquisition methods of feature vector in technical ability specifically include:
Feature vector of the worker in technical ability is obtained based on the technical ability of being good in worker data, it is a certain that use 0 indicates that worker does not have It is good at technical ability, use 1 indicates that worker has and a certain is good at technical ability, worker uiFeature vector in technical ability may be expressed as:
Ki=(k1, i, k2, i, k3, i..., kY, i)；
The worker specifically includes in the acquisition methods of intellectual feature vector:
Task category in task bid record based on worker and task data obtains the task category that worker participates in task；
Worker is obtained in intellectual feature vector based on the task category that worker participates in task；
Indicate that worker does not have a certain knowledge with 0, use 1 indicates that worker has a certain knowledge, worker uiIn intellectual feature vector It may be expressed as:
Si=(s1, i, s2, i, s3, i..., sX, i)；
The acquisition methods of feature vector of the worker in self-concept specifically include:
Worker data obtains task averagely participation amount after task based access control data and screening；
Retail shop in task based access control averagely participation amount and worker data improves rate and obtains feature vector of the worker in self-concept, Worker uiFeature vector in self-concept may be expressed as:
Ai=(pi, ri)
Wherein:
piFor worker uiTask averagely participation amount；
riFor worker uiRetail shop improve rate；
The acquisition methods of feature vector of the worker in speciality specifically include:
Based on obtaining worker in the speciality with grade, positive rating, speed, quality, attitude and collection number in worker data Feature vector, worker uiFeature vector in speciality may be expressed as:
Hi=(gi, ci, di, qi, ai, ni)
Wherein:
giFor worker uiGrade；
ciFor worker uiPositive rating；
diFor worker uiSpeed；
qiFor worker uiQuality；
aiFor worker uiAttitude；
niFor worker uiCollection number；
The acquisition methods of feature vector of the worker in motivation specifically include:
Feature vector of the worker in motivation, worker u are obtained based on the acceptance of the bid total degree in worker data and total amount of tradingiIn Feature vector in motivation may be expressed as:
Mi=(ti, bi)
Wherein:
tiFor worker uiAcceptance of the bid total degree；
biFor worker uiTransaction total amount.
9. a kind of crowdsourcing task recommendation system based on worker's capability comparison, which is characterized in that the system comprises computer, institutes Stating computer includes:
At least one storage unit；
At least one processing unit；
Wherein, be stored at least one instruction at least one described storage unit, at least one instruction by it is described at least One processing unit is loaded and is executed to perform the steps of
Task data and worker data on S1, acquisition crowdsourcing platform, are obtained based on the task data and the worker data The task bid record of worker, the task bid record based on worker screen worker data；
S2, task based access control data and screening after worker data obtain fusion qualified quality model worker's ability feature to Scale；
S3, similar worker similar to target worker's ability is obtained based on described eigenvector table；
S4, the task bid record based on similar worker generate preliminary task recommendation list；
S5, the task bid record based on generation preliminary the task recommendation list and target worker, generate final task and recommend List, and give the crowdsourcing task recommendation in the final task recommendation list to target worker.
CN201910723106.7A 2019-08-06 2019-08-06 Crowdsourcing task recommendation method and system based on worker's capability comparison Pending CN110443502A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201910723106.7A CN110443502A (en) 2019-08-06 2019-08-06 Crowdsourcing task recommendation method and system based on worker's capability comparison

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201910723106.7A CN110443502A (en) 2019-08-06 2019-08-06 Crowdsourcing task recommendation method and system based on worker's capability comparison

Publications (1)

Publication Number Publication Date
CN110443502A true CN110443502A (en) 2019-11-12

Family

ID=68433551

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201910723106.7A Pending CN110443502A (en) 2019-08-06 2019-08-06 Crowdsourcing task recommendation method and system based on worker's capability comparison

Country Status (1)

Country Link
CN (1) CN110443502A (en)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111626616A (en) * 2020-05-27 2020-09-04 深圳莫比嗨客数据智能科技有限公司 Crowdsourcing task recommendation method
CN111784308A (en) * 2020-07-16 2020-10-16 武汉空心科技有限公司 Human resource cost evaluation and pricing model training method based on working platform
CN112308432A (en) * 2020-11-03 2021-02-02 重庆理工大学 Crowdsourcing task recommendation method based on category and capability evaluation

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105930352A (en) * 2016-04-05 2016-09-07 扬州大学 Crowdsourcing task oriented exploratory search method
US20170046794A1 (en) * 2015-08-11 2017-02-16 Accenture Global Services Limited System for sourcing talent utilizing crowdsourcing
CN106776941A (en) * 2016-12-02 2017-05-31 济南浪潮高新科技投资发展有限公司 A kind of method of the effective solutionist of recommendation based on mass-rent pattern
CN108363804A (en) * 2018-03-01 2018-08-03 浙江工业大学 Partial model Weighted Fusion Top-N films based on user clustering recommend method
CN108804319A (en) * 2018-05-29 2018-11-13 西北工业大学 A kind of recommendation method for improving Top-k crowdsourcing test platform tasks
CN109241415A (en) * 2018-08-20 2019-01-18 平安科技（深圳）有限公司 Item recommendation method, device, computer equipment and storage medium
TWM577156U (en) * 2018-11-22 2019-04-21 全球華人股份有限公司 Competency analysis system for discipline cluster

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20170046794A1 (en) * 2015-08-11 2017-02-16 Accenture Global Services Limited System for sourcing talent utilizing crowdsourcing
CN105930352A (en) * 2016-04-05 2016-09-07 扬州大学 Crowdsourcing task oriented exploratory search method
CN106776941A (en) * 2016-12-02 2017-05-31 济南浪潮高新科技投资发展有限公司 A kind of method of the effective solutionist of recommendation based on mass-rent pattern
CN108363804A (en) * 2018-03-01 2018-08-03 浙江工业大学 Partial model Weighted Fusion Top-N films based on user clustering recommend method
CN108804319A (en) * 2018-05-29 2018-11-13 西北工业大学 A kind of recommendation method for improving Top-k crowdsourcing test platform tasks
CN109241415A (en) * 2018-08-20 2019-01-18 平安科技（深圳）有限公司 Item recommendation method, device, computer equipment and storage medium
TWM577156U (en) * 2018-11-22 2019-04-21 全球華人股份有限公司 Competency analysis system for discipline cluster

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
仲秋雁 等: "考虑用户兴趣和能力的众包任务推荐方法", 《系统工程理论与实践》 *
彭剑锋 主编: "《人力资源管理概论（第三版）》", 30 November 2018, 复旦大学出版社 *
王小妮 著: "《数据挖掘技术》", 31 August 2014, 北京航空航天大学出版社 *

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111626616A (en) * 2020-05-27 2020-09-04 深圳莫比嗨客数据智能科技有限公司 Crowdsourcing task recommendation method
CN111784308A (en) * 2020-07-16 2020-10-16 武汉空心科技有限公司 Human resource cost evaluation and pricing model training method based on working platform
CN112308432A (en) * 2020-11-03 2021-02-02 重庆理工大学 Crowdsourcing task recommendation method based on category and capability evaluation

Similar Documents

Publication Publication Date Title
Jarek et al. Marketing and artificial intelligence.
Yuen et al. Task matching in crowdsourcing
Yang et al. Question recommendation with constraints for massive open online courses
Kabir et al. Comparative Analysis Of TOPSIS and FUZZY TOPSIS for the Evaluation of Travel Website Service Quality.
Camp A bible for benchmarking, by Xerox
Huang Designing utility-based recommender systems for e-commerce: Evaluation of preference-elicitation methods
Fu et al. Explaining global box-office tastes in Hollywood films: Homogenization of national audiences’ movie selections
Chiu et al. Performance evaluation criteria for personal trainers: An analytical hierarchy process approach
US20110196716A1 (en) Lead qualification based on contact relationships and customer experience
CN110443502A (en) Crowdsourcing task recommendation method and system based on worker's capability comparison
Sage Entrepreneurship as an economic development strategy
Pawłowski et al. B2B customers buying behavior
Wang et al. Exploring influencing factors of offline knowledge service transactions on an online-to-offline knowledge-sharing economy platform
Einstein Nothing for money and your work for free: Internships and the marketing of higher education
ASLAN et al. Collaboration of business intelligence and cloud computing and selecting the best cloud business intelligence solution
US20140222700A1 (en) Predictive pairing and/or matching systems, apparatus, and methods
Chan-Tien et al. The conceptual framework for applying digital community marketing and marketing practices into educational relationship marketing model of private technical high school in Taiwan
CN106708939A (en) Target person scoring and pushing methods, apparatuses and systems
Wu et al. The optimal relationship between buyer and seller obtained using TOPSIS method
Lim et al. Maximizing bidder's profit in online auctions using grey system theory's predictor agent
CN115344767A (en) Supplier evaluation method based on network data
Siraj et al. Pre-selection of recruitment candidates using case based reasoning
Ritov et al. Transaction frame determines preferences: Valuation of labor by employee and contractor
Dikow et al. Improving the accuracy of Business-to-Business (B2B) reputation systems through rater expertise prediction
KR20180025674A (en) SW Market Information Service System

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
RJ01 Rejection of invention patent application after publication
RJ01 Rejection of invention patent application after publication

Application publication date: 20191112