공공 IT 산업 기술동향 및 참여기업 연결망 분석

This study analyzed the technology trends and structural characteristics of Korean public IT industry, based on contracts data, bidding data, and request for proposal (RFP) documents of public IT projects from 2013 to 2017. A Latent Dirichlet Allocation (LDA) based topic modeling method and Social Network Analysis (SNA) techniques have been utilized. Three different types of networks have been identified: topic-based networks, collaborative networks and competitive networks. By using topic modeling, relations between topics and companies have been identified based on the contract and RFP data. Then, topic-based networks have been established by similarities of topics between companies. Collaborative and competitive networks were constructed by analyzing the contract and bidding data of public IT projects. The networks showed that the public IT industry has been mainly reorganized with mid-sized firms since 2013. Collaborative and competitive networks showed a similar network topology; they both consist of one gigantic component, a few fragmented, small components and isolates. It is also interesting to note that most of collaborative and competitive relations were weak ties, which means only a small fraction of companies either collaborate or compete several times during the period.