Abstract
The technological keywords extracted from patent documents have much information about a developed technology. We can understand the technological structure of a product by examining the results of patent analysis. So far, much research has been done on patent data analysis. The technological keywords of patent documents contain representative information on the developed technology. As such, the patent keyword is one of the most important factors in patent data analysis. In this paper, we propose a patent data analysis model combining a integer valued time series model and copula direction dependence for integer valued patent keyword analysis over time. Most patent keywords are frequency values and keywords often change over time. However, the existing patent keywords analysis works do not account for two major factors: integer value and time. For modeling integer valued keyword data with time factor, we use a copula directional dependence model based on marginal regression with a beta logit function and integer valued generalized autoregressive conditional heteroskedasticity model. Using the proposed model, we find technological trends and relations in the target technological domain. To illustrate the performance and implication of our paper, we carry out experiments using the patent documents applied and registered by Apple company. This study contributes to the effective planning for the research and development of technologies by utilizing the evolution of technology over time.
Highlights
The role of patent analysis is to analyze the patent documents related to a given technology using statistical methods or machine learning algorithms
In nextusing section, carried experiments to illustrate the validity of our research models using the Figure 2 shows the proposed patent keyword analysis process using time series and copula models
To analyze the patent keyword data over time, we proposed a Gaussian copula directional dependence by using the beta logit model with an integer-valued generalized autoregressive conditional heteroskedasticity (GARCH) model for marginal distributions
Summary
The role of patent analysis is to analyze the patent documents related to a given technology using statistical methods or machine learning algorithms. Many patent analysis studies using visualization have been published in various fields [2,4,5,6] In addition to this network visualization technique, Kim et al (2017) performed patent analysis using statistical methods [7]. They considered penalized regression models based on ridge and least absolute shrinkage and selection operator (LASSO). Kim and Jun (2017) proposed a time series model considering integer valued elements because the patent-keyword matrix has integer valued elements which use the frequency values of keywords in patents [12]. We propose a patent analysis method that performs integer valued time series analysis and statistical relation analysis of technologies at the same time.
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
