Patents of 12 oriental medicinal plants using worldwide intellectual property service and patent cooperation treaty of World Intellectual Property Organization issued from January 2006 until December 2017 in 10 country groups were collected (PCT, 2018; WIPS, 2018).

The top 12 plants based on production rankings in Korea were selected namely: (KOSIS, 2018) magnolia-vine (Schisandra chinensis), balloon flower (Platycodon grandifloras), Cynanchi wilfordii radix (Cynanchum wilfordii), Hasuo (Pleuropterus multiflorus), Tian ma (Gastrodia elata), Sheng di huang (Rehmannia glutinosa), Chinese liquorice (Glycyrrhiza uralensis), Chinese boxthorn (Lycium chinense), Chinese yam (Dioscorea polystachya), Milk vetch (Astragalus mongholicus), Korean angelica (Angelica gigas), and Korean blackberry (Rubus coreanus).

Meanwhile, the 10 country groups were China, Europe Union (EU), India, Indonesia, International patents (PCT), Japan, Korea, Taiwan, USA, and Vietnam. The EU collects patents from its 31 member states, including France, Germany and the United Kingdom, which are also members of the European Patent Treaty.

In order to remove noise from the collected patent data, the collected patents were verified in four steps. First, representative technical keywords related to the 12 selected plants were selected by performing a reference search. Second, a total of 9,182 patents were collected from national patent registries based on the selected keywords. Third, a total of 3,498 redundant patents were removed based on the application number and title. In addition, based on the title, abstract and contents of the patents, duplicate patents were manually removed using the country-specific patent searching systems in China (SIPO, 2018), India (IPI, 2018), Indonesia (MLHR, 2018), Japan (JPO, 2018), Korea (KIPRIS, 2018), Taiwan (TIPO, 2018), and Vietnam (NOIP, 2018). Lastly, the specification and demand scope of the patents were analyzed. Eight hundred seventeen patents that were not related to the 12 selected plants after consultation with medicinal plant experts were removed.

The four-step verification process left us with 4,867 effective patents. The effective patents were classified into five technology categories such as genome, manufacture, food, medicine, and cosmetics. When a patent was duplicated in two or more technology categories, it was classified into only one core category based on its importance.

The IP portfolio (IPP) is a collection of IPs owned by a single entity. IPP analysis can suggest strategies to reduce investment risk in the future (Lin et al., 2006).

An IPP model was established to evaluate the growth stage of technologies related to the 12 medicinal plants, which were classified into five categories: beginning, growth, maturity, decline, and recovery.

For each medicinal plant in each country group, an IPP value (IPP_t) was determined, where t has a value from I to V corresponding to five time periods: I (2006 - 2007), II (2008 - 2009), III (2010 - 2011), IV (2012 - 2013), and V (2014 - 2015). The IPP_t was calculated by dividing the number of patents existing in the registries by the number of applicants during each period interval. However, the IPP value for the period 2016 - 2017 was not computed because of closed patents within that period.

The IPP model was obtained from equation (1)

where (s) are the 12 medicinal plants, (t) is the period interval, P_j is the total number (m) of patents (j), and A_i is the total number (n) of applicants (i).

The IPE values were used to estimate the development potential and competitive advantage of the medicinal plants in the five technology categories (genome, manufacture, food, medicine, and cosmetics) based on strategic positioning (Grimaldi et al., 2015) and patent growth rate (KIPI, 2005).

The strategic positioning of each patent was represented by the patent share ratio which is the ratio of each technology category to the total number of patents. The patent growth rate was represented by the growth rate which is the annual increase ratio of patents in the each technology category. The IPE values (IPE_s) for each plant from 2006–2015, where s has a value from 1 to 12 corresponding to the 12 medicinal plants were computed using the equation (2)

where (s) are the 12 medicinal plants, (i) is each technology category (i.e., genome, manufacture, food, medicine, and cosmetics), TC_i is the number of patents owned by each technology category (i), TP is total number of patents, and AI_i is the annual increase ratio of patents owned by each technology category (i).

Using the IPE_s, five technology categories were positioned into four quadrants of a scatter plot: upper right, continuous application; upper left, increase in recent applications; lower left, infancy applications; and lower right, decrease in recent applications.

The level of technology development related to the 12 medicinal plants using a cites per patent (CPP) and patent family size (PFS) indices were assessed.

The CPP gives the number of citations for patents to estimate the effect that patents have on technology development. A high CPP value suggests that a core patent or original patent has a high quality level. However, if the technology was developed recently, the CPP value may be lower because of fewer citations regardless of the importance of the technology (KIPI, 2005). The CPP index is an indicator of which patents in the target group (or country) have had a significant impact on subsequent innovation activities. The CPP was calculated using equation (3)

where (s) are the 12 medicinal plants, P_i indicates the cited numbers of patent (i), and R_i indicates the registered numbers of patent (i).

The PFS shows the importance of patents in commercial markets (Choi et al., 2014). The “patent family” indicates that the same patent remains in one or more patent offices of the country and this factor is interpreted as the degree of effort put forth to commercialize patents (Harhoff et al., 2003). A high PFS value is presumed to indicate an important patent with a high market power (Hicks et al., 2001). PFS was obtained from the equation (4)

where (s) are the 12 medicinal plants, F_i is the number (m) of parent family owned by patent (i) in a given country, and P_i is the total number (n) of patent (i) in that country.

The patent impact index (PII) and the technology strength (TS) index was used to evaluate the technological competitiveness of the 12 medicinal plants. The PII gives the degree to which other patents cite a given patent after it is registered. A high PII value indicates high technological competitiveness 3 (Huang et al., 200). The PII is a relative importance index that considers the technical level of the other patents that cite a given patent. It is generally used to identify relative skill levels that cannot be identified by the CPP index (KIPI, 2005). The PII is obtained from equation (5)

where (s) are the 12 medicinal plants, SP_a is the number of forward citations of specific patent (a), SR_a is the number of registrations of the specific patent (a), TP_i is the number of forward citations of total patents (i), and TR_i is the number of registrations of total patents (i).

The TS index simultaneously evaluates the quality and quantity of each patent. A high TS value means that a patent has strong technical power. The TS index can give an unbiased estimate of the level of technology, because it uses both registered and cited patents (Chen et al., 2007). The TS index for the 12 medicinal plants was calculated using equation (6)

where (s) are the 12 medicinal plants, PII_s is the patent impact index for 12 medicinal plants, and TP_i is the total numbers (t) of patent (i).

In order to present the best direction for research and development in each country group, a multi-layer patent model was developed with three factors. The IPM model was obtained from equation (7)

where (s) corresponds to each of the 12 medicinal plants, and is the blank technology (B), entry barrier (E), and future growth (F) of patent category (i).

Although the definition and scope of B, E and F were partly previously reported, in this equation, the following components were still defined. B represents the patent share ratio (for a specific category field or country group) relative to the total number of patents. It indicates the existence of undeveloped technologies with few related patents and few major applicants. B was classified according to the number of patents in a given category registered in the official patent office. If the patent registration rate is low, the relevant technology is widely developed, thus there is a risk of failing to develop a new patent in the related field. E represents how many patents were registered in the technology category. Meanwhile F reflects the rate of increase in the number of patents in a given category. If the number of registered patents in a given category increased rapidly during the period 2014 - 2015, it is assumed that the potential for future growth is high. A distribution-free ordinal logistic regression was used to complete the model (Endresen and Janda, 2015). The three factors of the model (i.e., B, E, and F) were quantified using a Likert scale with five ordinal levels (Sullivan and Artino, 2013; Wu and Leung, 2017).

The points were assigned on the Likert scale based on a survey of 30 members of the patent committee of the Foundation of Agricultural Technology Commercialization (https://www.fact.or.kr/) and Rural Development Administration (http://www.rda.go.kr/) in Korea. Ordinal logistic regression analysis were conducted using the SAS/STAT 12.1 software (SAS Korea, Seoul, Korea).

Of the 4,867 effective patents selected through the verification process of the collected patents, Korea had the highest number of effective patents at 2,424 (49.8%), followed by China at 1,958 (40.2%), USA at 159 (3.2%), and Japan at 120 (2.5%). Of the 12 medicinal plants, magnolia-vine and balloon flower accounted for 71% of the total patents. The proportional representation of the five technology categories among all of the patents was 38% medicine, 32% food, 17% manufacture, 8% cosmetics, and 5% genome. China, Japan, Korea and USA accounted for 95% of the total patent applications which is largely attributable to the fact that the investigated plants are oriental medicinal plants produced and consumed mainly in Asia.

The number of patents registered per year increased until 2012 and then decreased in the latter period. The magnolia-vine and balloon flower had the highest number of registered patents in all of the periods, but recently there was an increase in the number of patents for Korean angelica (Table 1).

In order to evaluate the technology growth of each plant, the growth stage of technologies using IPP model based on the correlation between the number of existing patents and the number of patent applications were investigated. China and Korea are the leading countries in whole technology growth related to the 12 medicinal plants.

For example, the magnolia-vine plant is in the decline stage because the numbers of patents and applicants continually increased from the first time period to the fourth time period, but they subsequently decreased in the fifth time period [Fig. 1 (A)]. The growth pattern of the balloon flower is similarly in the decline stage [Fig. 1 (B)]. However, trend of remaining 10 plants was unclear because of fewer patents or too few applicants.

Fig. 1.  Landscape of technologies related to two medicinal plants based on the IPP model of magnolia-vine (A) and balloon flower (B).

The time periods were; I (2006 - 2007), II (2008 - 2009), III (2010 - 2011), IV (2012 - 2013), and V (2014 - 2015). x-axis; number of applicants. y-axis; number of patents. The arrow direction is the time flow of periods. The circle sizes indicate the relative correlation between numbers of patents and applicants while the circle color indicates each medicinal plant.

In each country group, IPP model indicates that all country groups are at the beginning stage of technology growth except for China, Japan, Korea and USA. China is in the maturity stage of technology growth; the numbers of patents and applicants continually increased from the first time to the fourth time period and subsequently declined in the fifth time period.

The changing of patents indicates that rapid technological progress occurred between the third and fourth time periods. Korea is in the growth stage of technology growth because the numbers of patents and applicants both increased in the all of the time periods. Japan and USA are in the decline stage since the numbers of patents and applicants steadily decreased in all the time periods.

In the 12 medicinal plants, IPE model indicated an increase in recent applications for cosmetics technologies, an infancy stage for genomics technologies, and continuous applications for medicine and food technologies.

For the magnolia-vine, there was an increase in applications for cosmetics technology and a decrease in applications for food and medicine technologies [Fig. 2 (A)]. On balloon flower, there was an increase in applications for genome technology and food technologies [Fig. 2 (B)]. However, the category is expected to have high potential for development due to an increase in the use of materials in natural products such as pharmaceuticals and cosmetics (Pan et al., 2011). In the other 10 plants, Tian ma and Milk vetch showed an increase in applications for food technology. The sheng di huang, Chinese liquorice, Chinese boxthorn showed an increase in applications for medicine technology.

Fig. 2.  IPE model landscape of the potential for technology development based on magnolia-vine (A) and balloon flower (B).

x-axis; share ratio of patent. y-axis; increasing ratio of patent. The red line is average value of share ratio of patent (x-axis) and increasing ratio of patent (y-axis). The red line indicates the quadrants, upper right; continuous application, upper left; increase in recent applications, lower left; infancy applications, lower right; decrease in recent applications.

In the 10 country groups, IPE model was not meaningful because of small numbers of patents except in magnolia-vine and balloon flower. However, China and Korea had the greatest numbers of patents and applicants in all of the technological categories. For the magnolia-vine, China had a high growth and market share in food category while Korea had an increasing applications in the cosmetics category. For the balloon flower, China and Korea were the most active in food technology development. The other country groups had no meaningful rise in technology categories because of the small numbers of patents.

In order to evaluate the technological level of the 12 plants, the CPP and PFS indices were used. The CPP index evaluates the quality level of technology development and technological influence while PFS index evaluates the market capability and power.

The magnolia-vine and balloon flower had a lot of core patents or original patents which together had both high level of technology and strong market capability. The other plants had an average technology level with low market capability (Fig. 3).

Fig. 3.  The graph shows the distribution of technological levels and market power based on the CPP and PFS indices for the 12 medicinal plants.

The magnolia-vine and balloon flower have level of technology and strong market capability. The 10 medicinal plants have an average technology level with low market capability. x-axis; patent family size (PFS), y-axis; cites per patent (CPP).

In order to evaluate the technological competitiveness of the 12 plants, the PII and the TS index were used (Table 2). The magnolia-vine and balloon flower showed increases in both indices. For the balloon flower, the TS index increased despite stagnation in the PII. This was possible since the balloon flower was more influenced by the quantitative number than the qualitative level.

PII considers only the qualitative level of technology, whereas the TS index reflects both the qualitative level and quantitative measures. The other plants require aggressive marketing because of the average quality of the technology along with low market capability. The Sheng di huang plant had the highest PII despite its low TS index, indicating that it had the highest-quality patents among the 12 medicinal plants. The Cynanchi wilfordii radix plant had the greatest increase in TS index despite a decrease in its PII and low number of patents meaning there is a recent and rapid increase in the number of patents although there is a decrease in the level of technology.

In order to present a patent development strategy, a newly designed IPM model based on three factors (i.e., blank technology, entry barrier, and future growth) was used.

Our model predicted the best technology category for future patent development. For magnolia-vine, cosmetics technology was predicted as the proposed field for patent applications and medicine technology as the high-risk field for investment because of little blank technology and a high entry barrier [Fig. 4 (A)].

Fig. 4.  Radar chart of the relative Likert scale [0 (lowest) to 5 (highest)] with three technical factors (i.e., blank technology, entry barrier, and future growth).

Blue circle; blank technology, Black square; future growth, Red triangle; entry barrier. (A) Evaluation of magnolia-vine based on three factors in country groups, (B) Graph plot of medicine technology category based on three factors in 12 medicinal plants.

The ordinal logistic regression analysis for magnolia-vine showed that two factors have statistically significant p - values: blank technology and entry barrier (p - value < 0.01 for each factor), although the future growth factor was not significant (p - value = 0.17). Although various technology categories are predicted the best field for patent applications, ordinal logistic regression of the three factors was not significant in other plants except magnolia-vine and balloon flower.

In the technology categories for all medicinal plants the comparison of 12 plants was not significant due to some medicinal plants having small number of patents. The best medicinal plant was screened based on each technology category using a low entry barrier with two high factors (i.e., blank technology and future growth).

Our model indicated that medicine, genome, and cosmetic categories had high possibility of growth for patent applications while food and manufacture categories had low possibility for future patent development. For medicine category, the Korean blackberry, Chinese boxthorn, and Korean angelica were predicted to be the most valuable medicinal plants while balloon flower and Chinese yam will have high risks in future patent development [Fig. 4 (B)].