Service Providers’ Competence Identification in Knowledge-Intensive Crowdsourcing Context

Competence analysis of service providers is of great importance for a knowledge-intensive crowdsourcing platform to examine service provider’(SPs) effectiveness and contribution and therefore to improve its management and operation efficiency and accuracy. As the information highway highly-developed, there are a huge amount of online crowdsourcing communities where talent workers can exchange experience and ideas, which enables competence analysis using text mining. In this paper, we developed a competence identification framework to analyze and recognize SPs’ competence in online knowledge-intensive crowdsourcing (KIC) context. We firstly crawled the experience sharing articles, which contained excellent SPs’ experiences and ideas about the efforts that should be made to be successful, from several Chinese online crowdsourcing communities and then text mining techniques were applied to analyze these unstructured texts. Each sentence in the corpora was tokenized into several words, after which the words were clustered as different topics using Latent Dirichlet Allocation (LDA) model based on their underlying semantics. Furthermore, based on the LDA outputs, we identified six clusters of crowdsourcing SPs’ competence and thus constructed the competence system on the basis of Spencer’s competence dictionary and human intervention. Finally, the descriptions of the competence system were presented.