Measuring Unequal Knowledge Distance by Network Embedding and Multiple Relationships.

Knowledge distance, representing the dissimilarity between different knowledge units, has been considered as an important dimension of recombination novelty and technological innovation. Previous measurements merely rely on the citation relationship and ignore their directions and weights. To fill this gap, this study proposes a new measurement which not only captures the unequal citation relationship but also integrates multiple information to depict knowledge distance. The results show that our method can accurately portray the knowledge distance in both scientific areas and technical fields. [ABSTRACT FROM AUTHOR]