Crowdsourcing high-efficiency assignment method fusing heredity and greedy algorithms

Abstract

Crowdsourcing serves as a novel cooperative work concept and is gradually evolved into a successful business mode, and the crowdsourcing mode can be integrated into various social life and work of people. The crowdsourcing users earn corresponding rewards for themselves while solving the tasks, the free and flexible working mode is loved by more users and publishers, more and more tasks and solvers are gathered along with an open platform, and the crowdsourcing users are more and more interested in the open platform. The biggest challenge faced by crowdsourcing is how to accurately and efficiently assign tasks to appropriate task solvers. Based on constraint conditions in a crowdsourcing mode, a crowdsourcing task assignment method integrating a genetic algorithm and a greedy strategy is provided, and tasks with different requirements are accurately and efficiently matched with appropriate task solvers; by utilizing experimental analysis and performance comparison, it is verified thatthe method can completely achieve the purpose of accurate and efficient assignment in crowdsourcing problems.

Description

technical field [0001] The invention relates to a crowdsourcing high-efficiency distribution method, in particular to a crowdsourcing high-efficiency distribution method combining genetic and greedy algorithms, and belongs to the technical field of crowdsourcing mode distribution methods. Background technique [0002] With the rapid development of the Internet, more and more forms of demand and more and more transaction methods can be realized through the ubiquitous Internet. It is in this environment that crowdsourcing was proposed and received more and more attention. Highly concerned by users and enterprises. Crowdsourcing can be understood as group wisdom, which is used to describe a new business model. It refers to a company or enterprise that voluntarily publishes tasks previously performed by its own employees on the Internet. In other words, the company The problems and difficult problems to be solved are published on the Internet in a paid or unpaid way, expecting ...

AI Technical Summary

Problems solved by technology

[0009] Second, the task assignment of crowdsourcing platforms in practical applications will be subject to various constraints. It is necessary to explore the influence of multiple attributes and the combination of various situations. This is a typical NP-C problem. The existing technology uses pure mathematical analysis It is difficult to obtain the optimal solution of crowdsourcing assignment with a single algorithm or a single algorithm. There is a lack of a very effective method to solve the crowdsourcing model assignment problem by using group search. The coverage of existing technologies is small, it is difficult to find the global optimal solution, and it is easy to fall into a local optimal situation. ; The search method in the prior art starts from a single point to search and calculate. For the search space of multi-peak distribution, it is often limited to a single-peak optimal solution, with almost no parallel capability, and the risk of falling into a local optimal solution is very high. Poor performance in the process of global search optimization;

[0010] Third, the genetic algorithm in the prior art is completely random in generating the initial solution space for the target problem, making it difficult for the algorithm to quickly find the optimal solution, which increases the pressure on the subsequent iterative selection process. In the process of solving the assignment problem The local optimal solution cannot be obtained, the optimal solution of the genetic algorithm has no basic guarantee, and the subsequent operation of the genetic algorithm lacks corresponding processing, the diversity of the solution set is poor, the search range of the solution space is small, and the possibility of finding the global optimal solution lower;

[0011] Fourth, the existing crowdsourcing assignment system cannot realize the optimal assignment of crowdsourcing tasks based on the constraints between element attributes, and cannot assign the most suitable tasks to the most suitable people, and cannot achieve the optimal use of human resources

Crowdsourcing applicants have published a large number of micro-tasks of different categories and difficulties on the platform. Each crowdsourcing participant needs to browse these tasks and choose to perform the tasks that suit him. Crowdsourcing participants need to spend a lot of time looking for suitable tasks. The time spent on finding suitable tasks is often more than the execution time, which significantly reduces the efficiency and interest of crowdsourcing participants. The lack of efficient and accurate task assignment helps crowdsourcing A method for package participants to quickly and accurately select suitable tasks

Images