Fraud Risk Control And Intellectual Property Protection In Open Innovation ——A Study On The Behavior Of Sponsors In Crowdsourcing Contests

As one of the famous business models in knowledge economy,crowdsourcing is becoming an emerging model of open innovation.Crowdsourcing expands the source of creativity and innovation of enterprises,optimizes social resources,and benefits sponsors,platforms and workers.Research topics about crowdsourcing have aroused the wide attention of the academic circles.The stakeholders in crowdsourcing mode are connected because they complete the same task.This kind of short-term relationship contract lacks the stability of cooperation,which makes the sponsors and workers benefit and have risks at the same time.For example,the sponsor needs to have a detailed and complete understanding of all solutions provided by workers in advance before finally choosing the best one,a situation that provides an opportunity for sponsors to steal the solution without rewarding workers.This may lead to excessive competition among the crowdsourcing workers and the embezzlement risk of their intellectual property.Currently,manual detection of fraudulent sponsors by workers and platforms is time-consuming and laborious.There are few in-depth empirical studies on this kind of fraud as it relates to intellectual property(IP),for which the effective detection methods based on verbal cues identified in previous literature may lose their effectiveness,while there is a lack of application of computer technology in this field.Therefore,it is very challenging to automatically and accurately detect the fraudulent behaviors of sponsors.To address this challenge,this paper focuses on deepening the discussion of the value of various types of information cues for deception in the IP-oriented crowdsourcing market.Meanwhile,based on machine learning method,an effective fraud early warning system is constructed to automatically track and monitor crowdsourcing projects,and realize the whole process control and early warning of risk factors.Specifically,first,based on the traditional fraud theories,this study analyzes the fraudulent behavior of sponsors in crowdsourcing contests,and considering the quantification of structured and unstructured information,various machine learning methods are utilized to examine the effectiveness and value of various types of detection cues.Second,based on the above exploration of effectiveness,this study proposes a phased fraud monitoring and detection framework from the process perspective.Final,in the open innovation platform,the binary attribute,that is,the relationship network,may be more effective at identifying the potential behavior of users than a single source attribute.Thus,this study explores the effectiveness of social network analysis in the crowdsourcing sponsor fraud detection,to examine whether the relevant indicators of social network can effectively distinguish fraudulent sponsors,and ultimately improve the performance of the machine learning model.In fact,through the comparative analysis of verbal information and nonverbal information in static and dynamic contexts,this study confirms the effectiveness of traditional linguistic cues in IP-oriented online fraud detection,redefines the scope of online nonverbal cues and expands the theoretical boundary of fraud detection theory.In terms of different stages of the crowdsourcing contest,the fraud detection model and early warning framework not only enrich the application of process perspective(I-P-O)and social network analysis(SNA)theory,but also deepen the understanding of similar online activities.Taking the fraudulent behavior of crowdsourcing sponsors as the research object is not only to pay attention to the creative power of the public,but also an effective way to solve the problem that the public's enthusiasm is frustrated in the process of open innovation.Developing an intelligent fraud detection mechanism to reduce information overload and dispute solution costs,which helps to reduce the risk of crowdsourcing fraud and better build a fair platform environment.It is an important link to realize the protection of intellectual property rights in open innovation,as well as to maintain diversified innovation sources and sustainable and healthy development of crowdsourcing mode and open innovation.