The Theory Analysis And Application Of Security Mechanism Of Group Intelligence Software Development Based On Blockchain

The explosive growth and wide accessibility of the Internet have led to a surge in research activities related to swarm intelligence.Among them,software development methods based on Internet swarm intelligence have also been rapidly developed.Compared with traditional software development methods,the swarm intelligence software development model can maximize the use of the wisdom of developers distributed on the Internet,and complete complex software development tasks through their competition and cooperation,which greatly improves the quality and efficiency of software development.However,developers everywhere on the Internet cannot reach an effective trust consensus with the participants on the other side of the network.They usually regard the centralized service platform as an unconditional trust institution,and all transaction records and personal privacy data are stored on the platform.On the other hand,once the platform is attacked,it will cause service inaccessibility or user privacy leakage.In addition,the existing crowdsourcing platform lacks a good reputation system,most of which only consider the transaction amount and quantity,such as ignoring the contribution made by the developer reviewing code and the malicious behavior of users.Therefore,it lacks a good multidimensional consideration of user reputation evaluation indicators.Moreover,due to the transparent characteristics of the blockchain,the user's transaction information is exposed on the blockchain,and any node can query to associate and analyze the user's private data.Therefore,it is very important to design trustworthy,group-intelligent software development security mechanisms that have a good reputation system and can protect user privacy.In view of the above problems,the research work of this thesis can be divided into three aspects,each of which is as follows:(1)Design a decentralized group intelligence software development credible mechanism based on blockchain technology.The group intelligence software development process is divided into multiple sub-processes,and the metadata of each sub-process is published on the blockchain in combination with the hash algorithm,and each process is automatically executed through a smart contract.Developers,publishers,and platforms will act as nodes of the blockchain network to interact with smart contracts,and all transactions will be completed through the blockchain,without the need for a group intelligence software development platform as a trust center.The experimental results show that the mechanism can effectively solve the trust problem of the participants in the development of group intelligence software and the system has exceptional performance.(2)By examining the data of developers participating in tasks on the real group intelligence platform,a multi-dimensional consideration of the developer reputation evaluation system is proposed.Among them,the indicators to measure the credibility of a developer include the initial registration dimension,the transaction dimension for the developer to complete the development task,the developer's completion task to obtain the publisher's evaluation feedback dimension,and the developer's participation in the code review to make a contribution dimension,and taking into account the time factor and the impact of user malicious behavior.The results of the experiment show that the evaluation system can effectively and accurately reflect the credibility status generated by the various aspects of the developer's behavior.(3)Privacy protection based on homomorphic encryption.After the developer completes the software development task,the amount transaction will be performed,all of the amount transaction records will be stored in the blockchain in plaintext,and the homomorphic encryption technology will be used to homomorphically encrypt the user's transaction amount.Other verification nodes can verify the data in the ciphertext and get the result without knowing the specific amount of money so that user privacy is available and not visible.The experimental results show that this method can effectively protect user privacy information and has superior system performance.