Integer-valued GARCH processes for Apple technology analysis
DOI | https://doi.org/10.1108/IMDS-01-2017-0023 |
Pages | 2381-2399 |
Published date | 04 December 2017 |
Date | 04 December 2017 |
Author | Jong-Min Kim,Sunghae Jun |
Subject Matter | Information & knowledge management,Information systems,Data management systems,Knowledge management,Knowledge sharing,Management science & operations,Supply chain management,Supply chain information systems,Logistics,Quality management/systems |
Integer-valued GARCH processes
for Apple technology analysis
Jong-Min Kim
Department of Statistics, University of Minnesota at Morris,
Morris, Minnesota, USA, and
Sunghae Jun
Department of Statistics, Cheongju University, Cheongju, Republic of Korea and
Division of Industrial Management Engineering,
Korea University, Seoul, Republic of Korea
Abstract
Purpose –The keywordsfrom patent documents contain a lot of information of technology.If we analyze the
time series of keywords, we will be able to understand even more about technological evolution.
The previousresearches of time series processesin patent analysis were based on timeseries regression or the
Box-Jenkins methodology. The methods dealt with continuous time series data. But the keyword time series
data inpatent analysis are not continuous,they are frequency integervalues. So we need a new methodologyfor
integer-valuedtime series model. Thepurpose of this paper is to proposemodeling of integer-valuedtime series
for patent analysis.
Design/methodology/approach –For modeling frequency data of keywords, the authors used
integer-valued generalized autoregressive conditional heteroskedasticity model with Poisson and negative
binomial distributions. Using the proposed models, the authors forecast the future trends of target keywords
of Apple in order to know the future technology of Apple.
Findings –The authors carry out a case study to illustrate how the methodology can be applied to real
problem. In this paper, the authors collect the patent documents issued by Apple, and analyze them to find the
technological trend of Apple company. From the results of Apple case study, the authors can find which
technological keywords are more important or critical in the entire structure of Apple’s technologies.
Practical implications –This paper contributes to the research and development planning for producing
new products. The authors can develop and launch the innovative products to improve the technological
competition of a company through complete understanding of the technological keyword trends.
Originality/value –The retrieved patent documents from the patent databases are not suitable
for statistical analysis. So, the authors have to transform the documents into structured data suitable for
statistics. In general, the structured data are a matrix consisting of patent (row) and keyword (column), and its
element is an occurred frequency of a keyword in each patent. The data type is not continuous but discrete.
However, in most researches, they were analyzed by statistical methods for continuous data. In this paper,
the authors build a statistical model based on discrete data.
Keywords Patent analysis, Apple keywords, Integer-values time series model,
Poisson and negative binomial distributions
Paper type Research paper
1. Introduction
Many case studies on Apple’s technological innovation have been fulfilled in many
academic and industrial fields (Funk, 2011; Arruda-Filho et al., 2010; Arruda-Filho and
Lennon, 2011; West and Mace, 2010; Halal, 2013). The research studies focused on Apple’s
technological evolution. Apple is one of the global innovative companies leading the
smartphone market (Nam et al., 2015; Hung et al., 2013; Wonglimpiyarat, 2005). A number of
business schools have studied on the technological development and innovation of Apple
for other companies’technological innovations. Many companies have pursuit the
technological innovative strategies of Apple. To know the technological innovation of
Apple, we have to analyze Apple’s technologies. Apple’s patents contain more information
of developed technologies of Apple than other resources such as papers or articles because
the patent system protects the exclusive right of researched and developed technologies for
Industrial Management & Data
Systems
Vol. 117 No. 10, 2017
pp. 2381-2399
© Emerald PublishingLimited
0263-5577
DOI 10.1108/IMDS-01-2017-0023
Received 20 January 2017
Revised 19 March 2017
Accepted 31 March 2017
The current issue and full text archive of this journal is available on Emerald Insight at:
www.emeraldinsight.com/0263-5577.htm
2381
GARCH
processes
the inventors (Hunt et al., 2007). In this paper, we use entire patents applied by Apple for
Apple’s technology analysis. Jun and Park (2013) analyzed Apple’s patents using
statistical methods and the social network analysis for examining Apple’s technological
innovation ( Jun and Park, 2013). In addition, Kim and Jun (2015) proposed graphical
causal inference and the copula regression model for the keyword analysis of Apple’s
patents (Kim and Jun, 2015). They used advanced statistical inference and visualization to
construct Apple’s technology map. In this paper, we use technological keywords of patent
data. The keywords from Apple’s patent documents also contain diverse and complete
information of Apple’s technology. In addition, if we analyze the time series data of the
keywords, we will be able to understand even more about the technological evolution of
Apple. Guidolin and Guseo (2014) performed the seasonality modeling for Apple’s
innovative diffusion (Guidolin and Guseo, 2014).
It is meaningful to analyze patent data with the passage of time (Guidolin and Guseo,
2014; Hong et al., 2016; Lakka et al., 2013). In many studies on technology analysis of
company as well as Apple, diverse time series analysis models were used to understand
the technological trends of a company ( Jun and Uhm, 2010; Park and Jun, 2012; Jun, 2013;
Park et al., 2016). These research studies of time series processes in patent analysis were
based on time series regression or the Box-Jenkins methodology. The methods dealt with
continuous time series data. But the keyword time series data in patent analysis are not
continuous, they are integer values. So we need another methodology for the
integer-valued time series analysis for the patent keyword analysis. This paper discusses
the modeling of integer-valued time series with Apple’s keywords. For modeling count data of
the keywords, we use the integer-valued generalized autoregressive conditional
heteroskedasticity (INGARCH) model with Poisson distribution and negative binomial
distribution (Bollerslev, 1986; Christou and Fokianos, 2014; Davis and Wu, 2009; Engle, 1982;
Ferland et al., 2006; Fokianos and Fried, 2010, 2012). The results from numerical studies
indicate that the negative binomial INGARCH (INGARCH-NB) model performs better than the
Poisson INGARCH (INGARCH-Pois) model; and the INGARCH-NB model with covariate
keyword performs better than the INGARCH-NB model without covariate keyword.
In particular, using theApple’s keywords by text mining, w e show that the negative binomial
integer-valued autoregressive conditional heteroscedasticity (INARCH-NB) model has smaller
Akaike information criterion (AIC) value than any other models. So using the INARCH(1)-NB
model, we forecast our target Apple Keyword with other high correlated Apple Keyword in
order to know the future technology trend.
We organized our paper as follows. Section 2 introduces the way to extract keywords
from patent data. We propose count time series models for Apple’s keywords in Section 3.
In Section 4, we illustrate the examples with Apple’s keywords by text mining. Finally,
we show our conclusions in Section 5.
2. Keyword extraction form Apple’s patent documents
In this paper, we use technological keywords from retrieved Apple’s patent documents. First
of all, we retrieve entire patents applied by Apple in the world. We use the patent databases
of WIPS Corporation and the United States Patent and Trademark Office (WIPSON, 2014;
USPTO, 2015). Figure 1 shows the keyword extraction, structured data matrix, and
integer-valued time series modeling.
In this paper, we use R data language and its packages for data preprocessing and
analysis (Feinerer and Hornik, 2014; R Core Team, 2014). We build two structured data
matrices which are patent-keyword matrix (PKM) and year-keyword matrix (YKM). Using
the PKM, we make the correlation structure between Apple’s technological keywords, and
we perform time series analysis using the YKM. So, we construct integer-valued time series
modeling using the PKM and YKM.
2382
IMDS
117,10
To continue reading
Request your trial