Abstract: Technology roadmapping is a flexible tool to support emerging technology development path recognition and prediction, which has been widely used within countries, industries and enterprises. However, current technology roadmapping is mainly displayed by technical entities (keywords) as the core elements, which cannot effectively reflect the vein of technology development. In this paper, the subject-action-object (SAO) structures and natural language processing is used to help to construct technology roadmaps. First, we clear the domain key issue by natural language processing and experts’ knowledge. Then we extract SAO structure using Stanford parser software and analyze the technologies of solving key issues. After that, we display the relationships of the key issues and technology in the form of technology roadmaps. Finally, this paper starts an empirical study in the field of dye-sensitized solar cells (DSSC) with SCI-Expended data and experts’ knowledge, generates the technology roadmapping of improving the conversion efficiency of DSSC. The empirical shows that the method can provide an important reference for emerging technology innovation path recognition, and confirm the validity and applicability of the construction model for technology roadmapping.

摘要： 技术路线图作为一种新兴技术发展路径识别和预测的工具，正在被越来越多的国家、产业和企业认可接受。然而，当前技术路线图的研究和制定主要以定性方法为主，过分依赖专家经验；同时，技术路线图主要以技术实体（关键词）为核心显示要素，不能有效反映技术发展的脉络。因此，本论文从领域关键问题出发，运用文本挖掘技术和SAO结构分析方法，构建基于关键问题的核心技术方案群，并引入定量指标，以技术演化路线图的形式展示技术发展脉络，为新兴技术领域特定关键问题的未来技术方案研发及技术发展方向选择提供了重要参考。最后，文章以染料敏化太阳能电池 (DSSC) 技术为例开展实证研究，证实了该方法的科学性和有效性。

关键词: “主语-行为-宾语”(SAO), 技术路线图(TRM), 技术路径识别, 染料敏化太阳能电池(DSSC), Subject–Action–Object(SAO), Technology roadmaps (TRM), Technology path recognition, Dye-sensitized solar cells (DSSC)