A Flexible Crowdsourcing Ideation Method And Its Application

The Internet-based cyber-physical world has changed our world profoundly.The number of internet users increases and new data emerges,promoting the artificial intelligence research into the new era of artificial intelligence.Employing artificial intelligence to solve creative problems is an important trend in this new era.Crowdsourcing recruits crowds to solve problems that machine cannot solve yet in the form of an open call.Using crowdsourcing to solve creative problems could hopefully integrate people's creative capacity and machine's computational capacity,thus improving the creative capacity of artificial intelligence.IBM Corporation,BMW,and Centers for Disease Control and Prevention employed crowdsourcing to collect creative ideas;researchers recruited crowds to analyze protein structures,search particles,etc.However,crowdsourcing ideation was still unstable.Most crowdsourcing ideation methods published crowdsourcing tasks,and selected and refined ideas based on their quality.These methods focused on the quality of ideas and ignored the diversity of ideas,the further development potential of ideas,so they tended to be fixated on a few ideas in the early stage of crowdsourcing and restricted the crowds' capacity of creativity.Therefore,it was hard to produce high-quality ideas stably.This thesis studied the development of ideas during ideation process,and proposed a flexible crowdsourcing ideation method to improve the creative capacity of crowd intelligence.The flexible crowdsourcing ideation method motivated crowds to fulfill their creative potential and guided crowds to produce diverse ideas;then searched promising design directions from these ideas,selected and refined ideas that had high development potential,thus continually improving the quality of ideas.Specifically,this thesis conducted three following studies:(1)Proposed Creative Segment model to describe the ideation process.Creative Segment model integrated the ideation process of crowds during crowdsourcing,and supported the analyses of idea development process.(2)Refined the crowdsourcing tasks and guided crowds to produce diverse and high-quality ideas.This thesis analyzed participants' design ideation features and refined crowdsourcing tasks based on the source of ideas and the refinement process of ideas.The tasks encouraged crowds to inspire each other and improved the quality of ideas.(3)Refined the selection criteria of ideas.This thesis focused on the development potential of ideas rather than the quality of ideas,searching for the promising design directions during crowdsourcing.The thesis explored the relation among the idea distribution,the idea relation and the development results of ideas.An algorithm that calculated the development potential of ideas was proposed to predict the further development of ideas.The flexible crowdsourcing ideation method chose and refined ideas that had high development potentials,continually improving the quality of ideas.A flexible crowdsourcing ideation software was developed based on the flexible crowdsourcing ideation method.This software was then applied in three design ideation domains involving product design,human-computer interaction design,and graphic design.The results validated that flexible crowdsourcing ideation method kept exploring new design possibilities and improving the originality of ideas,and finally generated creative ideas.This thesis analyzed crowdsourcing ideation process,revealed the factors that influenced the development of ideas,calculated the development potential of ideas,and developed the promising directions accordingly,which improved the creative capacity of crowd intelligence.This thesis might also offer insights into other crowdsourcing methods and design methods.