Subject-Action-Object Structures Research Articles

Measuring Patent Similarity Based on Text Mining and Image Recognition

Patent application is one of the important ways to protect innovation achievements that have great commercial value for enterprises; it is the initial step for enterprises to set the business development track, as well as a powerful means to protect their core competitiveness. The emergence of a large amount of patent data makes the effective detection of patent data difficult, and patent infringement cases occur frequently. Manual measurement in patent detection is slow, costly, and subjective, and can only play an auxiliary role in measuring the validity of patents. Protecting the inventive achievements of patent holders and realizing more accurate and effective patent detection were the issues explored by academics. There are five main methods to measure patent similarity: clustering-based method, vector space model (VSM)-based method, subject–action–object (SAO) structure-based method, deep learning-based method, and patent structure-based method. To solve this problem, this paper proposes a calculation method to fuse the similarity of patent text and image. Firstly, the SAO structure extraction technique is used for the patent text to obtain the effective content of the text, and the SAO structure is compared for similarity; secondly, the patent image information is extracted and compared; finally, the patent similarity is obtained by fusing the two aspects of information. The feasibility and effectiveness of the scheme are proven by studying a large number of patent similarity cases in the field of mechanical structures.

Exploring Technology Opportunities Based on User Needs: Application of Opinion Mining and SAO Analysis

In recent years, as the importance of user innovation has become more important, enterprises have started to demand efficient and systematic user needs analysis for their products or services. Although opinion mining research based on a large number of online reviews available has been actively conducted recently, most existing studies that analyze user needs for technology development have provided inaccurate results due to the differences in vocabulary that exist between databases holding social data and patents. Motivated by this problem, we propose an approach to exploring technology opportunities that analyzes the subject-action-object (SAO) structures found in both patents and user reviews. To achieve this, we first carried out a sentiment analysis on user-review sentences linked to user needs, from this SAO analysis we extracted ample information related to technological structures contained in the reviews. Second, since in patent analysis, patent documents are structured in terms of the technological elements present, we extracted the technology’s SAO structures using their F-term code that contains multidimensional information available in this technology classification system. Finally, we vectorized the SAO structures derived from the reviews and patent documents using SAO2Vec before calculating the cosine similarity between SAOs in order to connect the reviews and patents. By applying the method proposed in this study to the automobile field, we present technological opportunities found in patents to address user needs found in automobile reviews.

METHOD OF DOMAIN ONTOLOGY AUTOMATED REPLENISHMENT FOR THE SUPPORT OF NEW TECHNICAL SOLUTIONS SYNTHESIS. PART 2

The article goes on to describe a solution to the problem of information support for the synthesis of new technical systems. The method consists in organizing the ontology of the subject area and filling it with data from the Russian-language patent array. An algorithm is presented for constructing a graph of elements of technical object structures from previously extracted primary semantic SAO (Subject–Action–Object) structures. The extracted bundles are preprocessed by searching for homogeneous sentence members and generating additional case forms. This is followed by linking the prepared SAO-objects into a single graph. The mechanism consists in sequentially transforming the subject and object actants into a set of anchor points from a common vocabulary of terms, followed by the memorization of the relation (predicate) for the identified points. The evaluation of data extraction by the system is carried out: the value of the F1 metric for a strict evaluation is 63 % and for a non-strict 79 %, respectively. The non-strict evaluation takes into account the correctness of the extraction of SAO root elements only. The extracted data is then converted into a subject domain ontology. The ontology scheme as a concept includes the structural elements of technical objects and the relationship between them, as well as supporting information on the invention. The initial content of the ontology is based on the processing of 11, 200 patent documents for inventions. The existing scheme already allows retrieving useful information about alternatives of structural components and communications between them. For example, searching for all elements of a structure in a given invention or tracking relationships. The results suggest that the proposed approach is promising. The authors see further research direction in improving the existing data extraction method and extending the ontology.

SAO2Vec: Development of an algorithm for embedding the subject-action-object (SAO) structure using Doc2Vec.

In natural-language processing, the subject–action–object (SAO) structure is used to convert unstructured textual data into structured textual data comprising subjects, actions, and objects. This structure is suitable for analyzing the key elements of technology, as well as the relationships between these elements. However, analysis using the existing SAO structure requires a substantial number of manual processes because this structure does not represent the context of the sentences. Thus, we introduce the concept of SAO2Vec, in which SAO is used to embed the vectors of sentences and documents, for use in text mining in the analysis of technical documents. First, the technical documents of interest are collected, and SAO structures are extracted from them. Then, sentence vectors are extracted through the Doc2Vec algorithm and are updated using word vectors in the SAO structure. Finally, SAO vectors are drawn using an updated sentence vector with the same SAO structure. In addition, document vectors are derived from the document’s SAO vectors. The results of an experiment in the Internet of things field indicate that the SAO2Vec method produces 3.1% better accuracy than the Doc2Vec method and 115.0% better accuracy than SAO frequency alone. This proves that the proposed SAO2Vec algorithm can be used to improve grouping and similarity analysis by including both the meanings and the contexts of technical elements.

Generic SAO Similarity Measure via Extended Sørensen-Dice Index

As an essential component of many Natural Language Processing applications, semantic similarity measure has been studied for decades. Recent research results indicate that the Subject-Action-Object (SAO) structure in sentences is more desirable for describing the technological information, and SAO-based similarity measure outperforms classical text-based ones. The typical approach in the literature to finding the similarity between two SAO structures relies on a term matching technique, which produces the similarity score by the Sorensen-Dice index, i.e., the proportion of the total number of matching terms. However, in this paper, we observe that the entities in the SAO structures usually have a small number of terms, which makes the currently acknowledged methods have a high recurrence rate and poor accuracy. To settle this issue, we extend the Sorensen-Dice index, and present a new unified framework for the SAO similarity measure that can give a higher discrimination. The effectiveness of our measure is evaluated on the basis of patent data sets in the Nano-Fertilizer field. The results show that our measure can significantly improve the accuracy than the currently acknowledged ones. The proposed measure has an excellent flexibility and robustness, and can be easily used for patent similarity measure. In addition, the extended Sorensen-Dice index is of independent interest, and has potential applications for other similarity measures.

Measuring patent similarity with SAO semantic analysis

Patents are not only an important aspect of intellectual property rights, but they are also one of the only ways to protect technological inventions. However, in recent years, the number of patents has been increasing dramatically and, as a result, both patent applicants and patent examiners are finding it more difficult to conduct the due diligence step of the patent registration process. Therefore, the lack of a quick and easy way to accurately measure patent similarity has become a significant obstacle to protecting intellectual property. Currently, there are three main ways to measure patent similarity: IPC code analysis, citation analysis, and keyword analysis. None of these approaches are able to fully reflect the semantics in a patent’s content. As an emerging methodology, subject–action–object (SAO) semantic analysis does reflect semantics, but most approaches treat each identified relationship as equally important, which does not necessarily provide an accurate measure of patent similarity. To offer this power to SAO analysis, this article introduces a new indicator called DWSAO as a reflection of the weight of each SAO semantic structure. Further, we present a semantic analysis framework that incorporates the DWSAO index for finding similar patents based on the weight of each SAO structure in the patent. A case study on the similarity of patents in the field of robotics was used to verify the reliability of the method. The results highlight the detailed meanings derived from the method, the accuracy of the outcomes, and the practical significance of using this approach.

Demand identification model of potential technology based on SAO structure semantic analysis: The case of new energy and energy saving fields

This study proposes an identification model based on subject–action–object (SAO) structure semantic analysis for the potential hotspots of technology demand to address the shortcomings of technology demand mining on the basis of word frequency statistical analysis. The SAO structure is extracted using Python tools to identify the potential hotspots of technology demand, the domain dictionary and professional corpus are introduced, and the clustering of technology demand is realized by applying Word2Vec and HowNet to calculate the semantic similarity among the SAO structures. The layout of the technology demand in the different stages of the technical lifecycle is divided by constructing a technology map. The proposed model is validated as an example of the network technology demand text of the new energy and energy saving fields. Therefore, the hotspots of technology demand are the technology of new energy vehicle motor and its control system, technology of energy efficient and technology of wind power, and the new energy vehicle technology is still in the research and development (R&D) stage. Moreover, solar energy products and production equipment are still in the technical application stage. This study provides an effective method for identifying potential technology demand and based on technology lifecycle to implement the layout and visualization of demand, which will make the decision support for guiding the direction of technology R&D, optimizing the allocation of science and technology resources, and promoting the effective docking of technology supply and demand.

Identifying R&D partners using SAO analysis: a case study of dye-sensitised solar cells

This paper proposes a systematic process to identify potential research and development (R&D) partners from a technological perspective based on subject-action-object (SAO) semantic analysis. Improvements to traditional methods are made by combining the SAO structure map and the collaboration network analysis. The SAO structure map reveals the technological development trends, organisations' research contributions and their research experiences in the field, which are the factors that indicate an organisation's R&D capabilities. Furthermore, we explore the organisation's collaboration statuses through collaborative network analysis and their collaborative publications, which make it easier to identify the organisation's sense of cooperation. Potential R&D partners are identified by examining the organisation's R&D capabilities and sense of cooperation. An exploratory study is conducted on dye-sensitised solar cells (DSSCs). The proposed method provides useful information for organisations (firms, institutions, universities, etc.) to identify potential R&D partners or make cooperation related policies.

Requirement-oriented core technological components’ identification based on SAO analysis

Technologies play an important role in the survival and development of enterprises. Understanding and monitoring the core technological components (e.g., technology process, operation method, function) of a technology is an important issue for researchers to develop R&D policy and manage product competitiveness. However, it is difficult to identify core technological components from a mass of terms, and we may experience some difficulties with describing complete technical details and understanding the terms-based results. This paper proposes a Subject-Action-Object (SAO)-based method, in which (1) a syntax-based approach is constructed to extract the SAO structures describing the function, relationship and operation in specified topics; (2) a systematic method is built to extract and screen technological components from SAOs; and (3) we propose a “relevance indicator” to calculate the relevance of the technological components to requirements, and finally identify core technological components based on this indicator. Based on the considerations for requirements and novelty, the core technological components identified have great market potential and can be useful in monitoring and forecasting new technologies. An empirical study of graphene is performed to demonstrate the proposed method. The resulting knowledge may hold interest for R&D management and corporate technology strategies in practice.

Combining SAO semantic analysis and morphology analysis to identify technology opportunities

Increasingly complex competitive environments drive corporations in almost all industries to conduct omnibearing innovation activities to enhance their technological innovation capability and international competitiveness. Against this background, we propose subject---action---object (SAO) based morphological analysis to identify technology opportunities by detecting prioritized combinations within the morphology matrix. SAO structures emphasize the key concepts with provision of diverse technology information based on semantic relationships. The combination of SAO semantic structures can support the establishment of matrix, which consists of two dimensions: compositions and properties of technology. Later, novel indicators are used to evaluate the subsequent technological feasibility of each new configuration under a customized analysis and prior combinations aided by a high score can be identified. We apply this method to the case of dye-sensitized solar cells (DSSCs) in patents documents. The approach holds promise to strengthen information support systems for commercial enterprises in technical innovation and market innovation activities. We believe the analysis can be adapted well to fit other technologies, especially in their emerging stage.

SAO Semantic Information Identification for Text Mining

A Subject-Action-Object (SAO) is a triple structure which can be used to both describe topics in detail and explore the relationship between them. SAO analysis has become popular in bibliometrics, however there are two challenges in the identification of SAO structures: low relevance of SAOs to domain topics; and synonyms in SAO. These problems make the identification of SAO greatly dependent upon domain experts, limiting the further usage of SAO and influencing further the mining of SAO characteristics. This paper proposes a parse tree-based SAO identification method that includes (1) a model to identify the core components (candidate terms for subject & object) of SAO structures, where term clumping processes and co-word analysis are involved; (2) a parse tree-based hierarchical SAO extraction model to divide entire SAO structures into a collection of simpler sub-tasks for separate subject, action, and object identification; and (3) an SAO weighting model to rank SAO structures for result selection. The proposed method is applied to publications in the Journal of Scientometrics (SCIM), to identify and rank significant SAO structures. Our experiment results demonstrate the validity and feasibility of the proposed method.

Subject–action–object-based morphology analysis for determining the direction of technological change

Morphology analysis, despite being a strong stimulus for the development of new alternatives, largely relies on domain experts and neglects the relationships between keywords in the construction of morphological structures. In addition, there are few systematic approaches to prioritize the morphological configurations. To address these issues, a hybrid approach is proposed, which enhances the performance of morphology analysis by combining it with subject–action–object (SAO) semantic analysis. Initially, a keyword co-occurrence patent set for subsequent SAO analysis is prepared based on keywords frequency vector analysis. Then, SAO structures are extracted and semantic analysis is performed to identify the relationships between keywords, which help to build morphological structures more objectively. In addition, a well-defined evaluation system that contains eight sub-indexes is proposed to evaluate the morphological configurations. Finally, to demonstrate and validate the proposed approach, the dye-sensitized solar cells technology is employed as the case study. Results indicate that the most promising combination we predict appears frequently in 2012–2014 and the distribution of it is also close to the fact in 2012–2014. Accordingly, the proposed method can be used to effectively determine the direction of technological change and to forecast technology innovation opportunities.

An SAO‐based text‐mining approach for technology roadmapping using patent information

A technology roadmap (TRM) links technologies with a company's strategic objectives and so supports acquisition of required technologies in advance of needs. It is a powerful tool for strategic planning and technology management. Because technology is changing rapidly and market competition is fierce, the role of a TRM is becoming increasingly important. To support the role of a TRM, many firms and governments that use roadmapping are becoming interested in reducing costs while retaining objectivity during TRM development. One suggestion to achieve these goals is to use the keyword‐based quantitative approach to creating a TRM, but the information provided by the approach is limited because of the characteristics of keyword information. To solve this limitation, this research uses the concept of ‘function’ to support quantitative analysis for developing a TRM. The concept of function can provide information on the uses and purposes of a technology. To represent a function, a subject–action–object structure is commonly used. The suggested approach allows research and development (R&D) managers to extend the views of product and technology during development of a TRM. In addition, by reducing the time required to develop a TRM, the proposed approach supports quick and accurate decision‐making by R&D managers.

A patent intelligence system for strategic technology planning

Patent intelligence—the transformation of content found in multiple patents into technical, business, and legal insight—is considered a key factor in gaining a competitive advantage in technologically competitive business environments. Although keyword-based patent intelligence tools are widely used due to their simplicity and ease of use, they are limited in that they cannot represent key technological concepts and inventive knowledge by relying only on the frequency of occurrence of defined keywords. As a remedy, this paper proposes a Subject–Action–Object (SAO)-based patent intelligence system. SAO structures that can be extracted from textual patent information are known as the expertise and inventive findings of the relevant patent. On the basis of semantic analysis of patent SAO structures, our proposed intelligence system constructs patent maps and patent networks. Building on the maps and networks, the system provides specific functionalities including identification of technology trends and significant patents, detection of novel technologies, and identification of potential infringement. This paper describes the architecture of our proposed patent intelligence system in detail, and illustrates the system’s functionalities using case studies. We anticipate that our proposed system will be incorporated into the technology planning process to assist experts in the formulation of technology strategies.

An SAO-based text mining approach to building a technology tree for technology planning

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 fact-oriented ontological approach to SAO-based function modeling of patents for implementing Function-based Technology Database

Function-Oriented Search (FOS) has been proposed as a tool for use in searching patent databases to find existing solutions to new problems. To implement FOS effectively, a well-structured Function-based Technology Database (FTDB) is required. An FTDB is a data repository of technology information represented as “function”. To implement an FTDB, four features should be addressed: continual data updating, limited area searching, function generalization, and semantics handling. In this paper, we consider these features to suggest a fact-oriented ontological approach to implementing an FTDB by Subject–Action–Object (SAO)-based function modeling of patents. The proposed approach uses fact-oriented ontology modeling of SAO structures extracted from patent documents, and implements an FTDB, which is an SAO-based patent retrieval system to support FOS. We also verify the feasibility of the proposed approach to by using it to conduct case studies of patent retrieval.

Identifying rapidly evolving technological trends for R&D planning using SAO-based semantic patent networks

Patents constitute an up-to-date source of competitive intelligence in technological development; thus, patent analysis has been a vital tool for identifying technological trends. Patent citation analysis is easy to use, but fundamentally has two main limitations: (1) new patents tend to be less cited than old ones and may miss citations to contemporary patents; (2) citation-based analysis cannot be used for patents in databases which do not require citations. Naturally, citation-based analysis tends to underestimate the importance of new patents and may not work in rapidly-evolving industries in which technology life-cycles are shortening and new inventions are increasingly patented world-wide. As a remedy, this paper proposes a patent network based on semantic patent analysis using subject-action-object (SAO) structures. SAO structures represent the explicit relationships among components used in a patent, and are considered to represent key concepts of the patent or the expertise of the inventor. Based on the internal similarities between patents, the patent network provides the up-to-date status of a given technology. Furthermore, this paper suggests new indices to identify the technological importance of patents, the characteristics of patent clusters, and the technological capabilities of competitors. The proposed method is illustrated using patents related to synthesis of carbon nanotubes. We expect that the proposed procedure and analysis will be incorporated into technology planning processes to assist experts such as researchers and R&D policy makers in rapidly-evolving industries.