روش متن کاوی مبتنی بر SAO برای ساخت درخت تکنولوژی برای برنامه ریزی فناوری

A technology tree (TechTree) is a branching diagram that expresses relationships among product components, technologies, or functions of a technology in a specific technology area. A TechTree identifies strategic core technologies and is a useful tool to support decision making in a given market environment for organizations with specified capabilities. However, existing TechTrees generally overemphasize qualitative and expert-dependent knowledge rather than incorporating quantitative and objective information. In addition, the traditional process of developing a TechTree requires vast amounts of information, which costs considerably in terms of time, and cannot provide integrated information from a variety of technological perspectives simultaneously. To remedy these problems, this research presents a text mining approach based on Subject–Action–Object (SAO) structures; this approach develops a TechTree by extracting and analyzing SAO structures from patent documents. The extracted SAO structures are categorized by similarities, and are identified by the type of technological implications. To demonstrate the feasibility of the proposed approach, we developed a TechTree regarding Proton Exchange Fuel Cell technology.

A technology tree (TechTree) is a branching diagram (tree structure) that represents relationships among technologies. Generally, a TechTree is used to represent the relationships among product components, technologies, or functions of a technology in a specific technology area. A TechTree is a useful tool for supporting a decision making process to identify strategic core technologies in a given market environment and organization capabilities. Furthermore, it helps to define element technologies based on identified core technologies. Due to these advantages, TechTrees are important decision-making tools for technology planning (Cheong, 2006), and are utilized in technology planning (Aude and Kahn, 1986, Choudhury and Fallah, 2009, Durand, 1992, Guglielmi et al., 2010, Visentin, 2008 and Yoon et al., 2008). The current process of developing a TechTree has the following limitations. The first is exclusive dependence on experts during the development process. Currently, most TechTree diagrams are created by reflecting the opinions of technology domain experts. However, developing a TechTree by this process has considerable costs in terms of time and human resources; and obtaining the participation of an expert can be difficult (Heiss & Jankowsky, 2001). Recently, new technologies have continuously appeared and replaced old technology. To establish strategic technology planning that considers the rapid technological changes, a TechTree should be constantly updated to reflect the changes. However, because of the current process, updating and managing TechTrees are difficult. The second limitation is that current TechTrees only represent limited technology information. A TechTree can represent various technology perspectives such as product taxonomy (Aude and Kahn, 1986, Cascini and Zini, 2008, Durand, 1992 and Yoon et al., 2008), technology taxonomy (Choudhury and Fallah, 2009, Durand, 1992, Guglielmi et al., 2010 and Visentin, 2008), and function taxonomy of technology (Cascini and Zini, 2008 and Cheong, 2006). However, the TechTree represents technology information from only the technology perspective for which the tree was developed. Another limitation is that existing TechTrees cannot represent subsidiary information related to technologies. In technology analysis, the subsidiary information of technology is as important as the technology itself. We called this ‘technology meta-information’. This information includes the inventor of the technology, the time at which the technology was developed, or where it was developed. By integrating these various perspectives of technology and technology meta-information, we can generate a TechTree that helps decision-making in technology planning more effectively than do existing TechTrees. To address these problems, we suggest using Subject–Action–Object (SAO)-based text mining techniques to develop TechTrees. This approach analyzes patent documents and exploits the analysis results to create a TechTree that represents technology perspectives. Patent information includes valuable up-to-date technological information and is continuously updated. Patents also include bibliometric information and this information can be utilized as technology meta-information. The proposed approach consists of a procedure for constructing source data from patents and a method of developing a TechTree using the data. The procedure uses Natural Language Processing (NLP) technology to extract SAO structures from patents, and text-mining techniques to analyze the SAO structures. The resulting information is used as the source data for development of a technology tree. The method develops a TechTree that represents various technology perspectives such as product taxonomy, technology taxonomy, and function taxonomy of technology. The suggested TechTree is represented with technology meta-information. The rest of paper is organized as follows. In Section 2 we describe work related to the proposed approach. In Section 3 we provide a detailed description of our approach. In Section 4 we illustrate the proposed framework by describing a case study of developing a TechTree that represents the technology used in Proton Exchange Membrane Fuel Cells (PEMFCs). In Section 5 we provide concluding remarks and directions for further study.

In spite of the importance of TechTrees as tools for technology planning, little research has been conducted on their development. This shortage has caused the limitation that dynamically updating and maintaining TechTrees is difficult; this is a serious problem because technology is advancing so quickly. To remedy these problems, we presented an SAO-based text mining approach to developing TechTrees. The proposed approach extracts SAO structures from patent documents, and analyzes them to develop a TechTree. Text-mining techniques are used to extract SAO structures, which are then categorized by similarity, and are identified by the type of technological implications that they represent. To demonstrate the feasibility of the proposed approach, we developed a case study for PEMFC technology. The proposed approach provides a quantitative method that uses patent information, rather than experts’ knowledge, and that considers technology meta-information when constructing the TechTree. This approach will be used as a useful tool to help researchers or R&D policy makers to conduct technology planning. Furthermore, we expect that it can be also used for strategic R&D planning by exploiting various types of technology meta-information. For example, in the merger and acquisition process the decision maker can utilize the method to identify technological advantages by comparing the TechTrees of different companies. To make the proposed approach more practical, three challenges must be addressed. First, a more efficient method extracting SAO structures should be developed. Knowledgist™ is almost the only tool available for extracting SAO structures. Although it is easy to use, it extracts many irrelevant SAO structures. This problem means that time must be spent removing irrelevant SAO structures and reduces the reliability of the TechTree constructed using the proposed approach. A better SAO extractor can guarantee a TechTree of better quality. Second, well-defined technology thesauri or ontologies should be constructed. In the proposed approach, we manually analyzed technological word phrases to identify their types. Well-defined thesauri would be more adequate for resolving identification of technology words. Finally, to support the development of a TechTree systematically, an analysis system should be constructed. Currently, we use Knowledgist™ to extract SAO structures and to develop a system of constructing a TechTree. However, the current system cannot fully support the overall process of constructing a TechTree. If an optimized system that includes a technological database can be developed, it would improve the practicality of the proposed approach.