This study characterizes and classifies the assignees of a technology field's patents through quantitatively determining their structural positions against a trajectory epitomizing the field's knowledge evolution. By considering that these patents’ citation network embodies a knowledge structure for the technology field, and assuming that a series of mainstream (MS) patents constitute the evolutionary trajectory, each non-MS patent is identified to be at one of the following positions: forward and backward reachable (FBR), backward reachable only (BRO), forward reachable only (FRO), and unreachable (UR), based on their reachability with the MS patents. The assignees are then associated with five positioning attributes, which are the shares of their patents at respective positions. With precise definitions using these quantitative attributes, assignees of the technology field are classified into exactly one of the distinctly positioned categories, namely trendsetters, contributors, absorbers, bystanders, and reinforcers, or one of the multiply positioned categories of mixed characteristics. These categories can be geometrically interpreted and the assignees’ positions can be visualized in a three-dimensional positioning space. This study then uses U.S. biochip patents and evolutionary trajectory derived by main path analysis (MPA) to observe how the proposed method works.

本研究通过定量确定技术领域专利的受让人相对于代表该领域知识演变的轨迹的结构位置来表征和分类技术领域的专利权人。通过考虑这些专利引网络体现了该技术领域的知识结构，并且假设一系列主流（MS）专利构成的进化轨迹，每个非MS专利被识别为在以下位置中的一个：向前和向后可达(FBR)、仅向后可达(BRO)、仅向前可达(FRO) 和不可达(UR)，基于它们与 MS 专利的可达性。然后，受让人与五个定位属性相关联，这些属性是他们在各自位置的专利份额。通过使用这些定量属性的精确定义，技术领域的受让人被准确地分类为定位明确的类别之一，即潮流引领者、贡献者、吸收者、旁观者和强化者，或混合特征的多重定位类别之一。这些类别可以进行几何解释，并且可以在三维定位空间中可视化受让人的位置. 然后，本研究使用美国生物芯片专利和通过主路径分析 (MPA) 得出的进化轨迹来观察所提出的方法是如何工作的。