How to use the big data to the technology planning: A data-driven technology roadmapping using ARM

Due to the explosive rise of data scale and scope in the business environment, the active incorporation of big data becomes an imperative and vital process in the technology roadmap. However, technology roadmapping still remains as a subjective and instinctive task conducted by some experts. Especially, the identification of relationship among different layers has been mainly dependent upon the intuitive judgment of experts. Some previous research, albeit infrequent, has attempted rather scientific approach but yet has been subject to limitations in that it remains as simply calculating the frequency of occurrence for keywords, which only provides the similarity. However, what is required to the technology roadmapping is to identify the dependency between layers. In response, this paper suggests an association rule mining (ARM)-based technology roadmap to provide both the affinity and dependency information between two layers. For this purpose, two types of indexes are measured using ARM: support and confidence. After measuring each rule, two types of maps are developed: affinity map and dependency map. For the intra-layer relationship, the affinity map is developed whereas the dependency map is constructed for the inter-layer relationship.