Exploration of a Concept Screening Method in a Crowdsourcing Environment
As an effective way to aggregate a crowd’s wisdom,crowdsourcing has attracted much research attention in recent years.Especially for product design and development,crowdsourcing shows huge potential for creativity and has been regarded as one important approach to acquire innovative concepts.However,there is still a challenge to make use of crowdsourcing in product design: how to review the large number of crowdsourcing concepts efficiently.To tackle this problem,a concept screening method is proposed in this article so as to i) improve the efficiency of evaluating crowdsourcing results; and ii) assist designers in identifying promising design candidates for further evaluation.Firstly,web mining technique is applied to extract textual information from Webs and tokenize text contents into word tags.Secondly,a concept similarity estimation process is designed to calculate the similarity between design requirements and crowdsourced concepts.During the estimation process,three situations are considered: 1) similarity caused by repeated tokens; 2) similarity caused by synonymous tokens; and 3) similarity in terms of design knowledge hierarchy.Finally,concepts are clustered based on their similarities,and the ones,which meet design requirements better,will be identified as promising candidates to be further reviewed by designers.To validate the proposed method,a pilot study on future PC design is presented.
Crowdsourcing web mining design knowledge hierarchy concept clustering
Danni CHANG Chun-Hsien CHEN
School of Mechanical & Aerospace Engineering, Nanyang Technological University,Singapore
国际会议
The 21st ISPE International Conference on Concurrent Engineering (2014国际并行工程学术会议)
北京
英文
861-870
2014-09-08（万方平台首次上网日期，不代表论文的发表时间）