Main path analysis considering citation structure and content: Case studies in different domains

Main path analysis (MPA) is an effective method widely accepted in science and technology for extracting knowledge diffusion paths. Traditional citation analysis assumes that all citations are treated equally. In contrast, this paper proposes a new MPA framework from the perspective of citation structure and content. Three indicators are considered to adjust edge weight: (1) Struc-tural similarity, (2) Topic similarity and (3) Sentiment analysis. This study takes the bullwhip effect and the Internet of Things domain as examples to verify the reliability and feasibility of improved MPA. The results show that the improved main path uncovers the knowledge trajec-tories appropriately, which has an ability to distinguish citations and detect important papers. This research enriches MPA theory and provides future research directions from perspective of citation structure and content.

Automated summary

Main Path Analysis (MPA) is an effective method for extracting knowledge diffusion paths. This paper proposes a new MPA framework based on citation structure and content. Three indicators are used to adjust edge weight: Structural similarity, Topic similarity, and Sentiment analysis. The study verifies the reliability and feasibility of improved MPA using the bullwhip effect and the Internet of Things domain as examples.