Topological Analysis of the Language Networks of Ancient Traditional Chinese Medicine Books

Zou, Qunsheng; Wang, Yinyan; Shu, Zixin; Yang, Kuo; Wang, Jingjing; Lu, Kezhi; Zhu, Qiang; Liu, Baoyan; Zhang, Runshun; Zhou, Xuezhong

doi:https://doi.org/10.1155/2020/8810016

Evidence-Based Complementary and Alternative Medicine

On this page

Abstract Introduction Materials and Methods Results Discussion Conclusion Data Availability Conflicts of Interest Acknowledgments References Copyright Related Articles

Special Issue

Innovative Computational Methodology for Complementary and Integrative Medicine Research

View this Special Issue

Research Article | Open Access

Volume 2020 | Article ID 8810016 | https://doi.org/10.1155/2020/8810016

Topological Analysis of the Language Networks of Ancient Traditional Chinese Medicine Books

Qunsheng Zou,^1,2Yinyan Wang,^1,2Zixin Shu,^1,2Kuo Yang,^1,2Jingjing Wang,^1,2Kezhi Lu,^1,2Qiang Zhu,^1,2Baoyan Liu,³Runshun Zhang,⁴and Xuezhong Zhou^1,2

Academic Editor: Zhaohui Liang

Received13 Sept 2020

Accepted24 Nov 2020

Published11 Dec 2020

Abstract

This study aims to explore the topological regularities of the character network of ancient traditional Chinese medicine (TCM) book. We applied the 2-gram model to construct language networks from ancient TCM books. Each text of the book was separated into sentences and a TCM book was generated as a directed network, in which nodes represent Chinese characters and links represent the sequential associations between Chinese characters in the sentences (the occurrence of identical sequential associations is considered as the weight of this link). We first calculated node degrees, average path lengths, and clustering coefficients of the book networks and explored the basic topological correlations between them. Then, we compared the similarity of network nodes to assess the specificity of TCM concepts in the network. In order to explore the relationship between TCM concepts, we screened TCM concepts and clustered them. Finally, we selected the binary groups whose weights are greater than 10 in Inner Canon of Huangdi (ICH, 黄帝内经) and Treatise on Cold Pathogenic Disease (TCPD, 伤寒论), hoping to find the core differences of these two ancient TCM books through them. We found that the degree distributions of ancient TCM book networks are consistent with power law distribution. Moreover, the average path lengths of book networks are much smaller than random networks of the same scale; clustering coefficients are higher, which means that ancient book networks have small-world patterns. In addition, the similar TCM concepts are displayed and linked closely, according to the results of cosine similarity comparison and clustering. Furthermore, the core words of Inner Canon of Huangdi and Treatise on Cold Pathogenic Diseases have essential differences, which might indicate the significant differences of language and conceptual patterns between theoretical and clinical books. This study adopts language network approach to investigate the basic conceptual characteristics of ancient TCM book networks, which proposes a useful method to identify the underlying conceptual meanings of particular concepts conceived in TCM theories and clinical operations.

1. Introduction

As a traditional medicine with medical theories and concepts mainly established thousands of years ago, TCM has abundance of high-value ancient books written or printed in the form of Chinese classical binding before 1912, which conceive important TCM theories and concepts and the clinical principle for disease diagnosis and treatment [1, 2]. Although many TCM antecessors performed significant theoretical investigations to digest the knowledge employed in these books, which has promoted the advances of contemporary clinical practical solutions for the managing of various complicated diseases in real-world clinical settings [3], it is particularly important to investigate the language characteristics of ancient TCM books, which will help understand the theoretical knowledge exactly expressed in those texts [4]. However, little research was conducted to understand the language regularities of the key TCM concepts (e.g., Yin, Yang, and Qi) in these ancient books using computational linguistics and complex network approaches [5].

Complex network has been developed as a mainstream approach for investigating the regularities in the fields with complex phenomena, such as social science, biological science, and linguistics [6–8]. For this approach, network or graph consisting of nodes and links is the form to represent the structures of the related systems. Due to the complex organization and interactions between various medical entities, complex network approaches have been used for exploring the rules of associations between herbs, symptoms, syndromes, and human meridians [9–12]. However, rare work was conducted on the analysis of language regularities of TCM books by complex network approaches [13].

In this paper, firstly, we constructed directed networks from the full texts of ancient TCM books. Then, we analyzed the statistical characteristics of the networks, identified the centrality patterns of core TCM concepts, and explored the similarities and differences between different ancient books. In addition, we demonstrated what the diversity of the concepts, such as “Qi (气)”, “Yin (阴)”, “Yang (阳)”, “Xie (泻)”, and “Li (痢)”, in Chinese medicine, and what special conceptual meanings they would have.

2. Materials and Methods

2.1. Dataset of 80 Ancient TCM Books

The data we used was derived from texts of 80 ancient Chinese medicine books (Table 1 shows a typical collection of 30 books), with an emphasis on the analysis of the books of Inner Canon of Huangdi (ICH) and Treatise on Cold Pathogenic Disease (TCPD). For example, ICH contains 189,984 characters and TCPD contains 43,331 characters. Data cleaning was performed to remove the characters in ancient texts except for Chinese and periods and separated the whole text into sentences. For example, ICH and TCPD, after cleaning the data of these two ancient books, we obtained 6,237 sentences and 1,366 sentences, respectively.

2.2. Language Network Construction Using 2-Gram Model

In the field of computational linguistics, n-gram is a widely used method to model natural language; particularly, an n-gram is a contiguous sequence of n items from a given sequence of text [14, 15]. Here, we used 2-gram model to obtain the sequential links of characters in ancient TCM books. Given a sentence, we would generate a directed path with characters as nodes and the sequential associations between them as links. When all the sentences of a given book were processed, we would obtain a weighted directed language network, in which the number of identical sequential associations is considered as the weight of the link. For example, the sentence “阴阳者, 天地之道也” in ICH can be processed to a directed path (Figure 1(a)) [16]. We have built the language networks for all the 80 ancient TCM books. In particular, the network of ICH contained 2,367 nodes and 35,502 directed links (see Figure 1(b)).

(a)

(b)

2.3. Basic Network Characteristics

The number of links connected to each node in TCM book networks, that is, the degree of the node [17]. We counted the degree of each node in the network and figured up the number of nodes with the same degree and attempted to find out whether the degree distributions of ancient book networks are consistent with the power law [18]. By calculating the average path length [19] and clustering coefficients [20] of networks, we judged whether these networks possess the small-world property. Thus,where is the average path length of graph , n is the number of nodes, and denotes the shortest distance between and . When cannot be reached from , . Moreover, clustering coefficients are acquired by

Consider as the local clustering coefficient of node i, where is the degree of node i, is a set of nodes which immediately connected with node i, E is defined as a set of edges in graph , and is the edge of nodes j and k. Then, all were summed and averaged to get the average clustering coefficient .

2.4. Homogeneity of the Centrality of Similar TCM Concepts in the Language Network

In TCM theories, there are hundreds of similar basic concepts (often in the form of a single character word), such as “Yin and Yang (阴阳)”, “the five elements (五行)”, and “the five internal organs (五脏)”, which are essential for TCM theories and clinical solutions [21]. We supposed that this kind of essentiality or importance of the concepts could be captured by vector representation of nodes in the directed language networks. In addition, we assumed that for those similar concepts, they would finally have similar centralities; that is, the similar concepts would display the same degree of centrality homogeneity compared with the random sets of concepts. The vector representation of each node in book network is calculated by the Node2Vec framework which learns low-dimensional representations for nodes in a graph [22]. To investigate the homogeneity phenomenon of these similar concepts, we proposed the following methods to differentiate between similar concepts and their random controls:where and are the low-dimensional representation of nodes, and are their components, and , is the dimension of vectors.

Determine as a set of similar concepts and as the set of another nodes in network. The similarity of concepts in groups will be defined by and similarity between in-group concepts and out-of-group concepts will be defined by , where and . If our hypothesis is correct, then is generally greater than .

2.5. t-Test on Similarity Sequences

Through the above method, we gained the difference between similarity of each basic concept and random concept. However, the current results were only relative to a single basic concept, which did not indicate that similar concepts are homogeneous to some extent. There is a way to solve this problem, named the Student t-test, which is often used to assess whether the means of two classes are statistically different from each other by calculating a ratio between the difference of two class means and the variability of the two classes [23, 24]. In this way, we can verify the homogeneity of basic concepts indirectly through the results of t-test. Whereupon, we combined results of each basic concept into similarity sequence M and results of random concepts into similarity sequence N. After performing t-test on these two sequences, if value is less than 0.05, we believe these basic concepts are similar to each other and consistent with homogeneity.

2.6. Identifying the Concept Clusters

It is well recognized that the complex networks like language network often hold a kind of community structures with some subnetworks involving dense links while sparse links outside those subnetworks. These subnetworks, which are considered as network clusters or communities, would deliver domain meaningful knowledge for further investigation. To detect the concept clusters or communities in the TCM language network, we applied the Fast Unfolding Algorithm (FUA) which was a well-known community detection method based on modularity [25] to detect the communities of a given network by

Think of Q as the modularity of the entire network, where represents the sum of the weights of all the edges in the network, is the weight between node i and node j, is the sum of the weights of the edges which connected to node i, and indicates the community which node i is assigned to. The value of is 0 or 1; when , it means node i and node j are in the same community; otherwise, node i and node j are not in the same community. Then, iteratively making the modularity reach the maximum value, the final clustering results were obtained.

3. Results

3.1. Basic Characteristics of TCM Language Network

It is observed that the degree distribution of ancient Chinese medicine nodes is consistent with the power law distribution [26, 27] (Figures 2(a) and 2(b)), which means that although most characters were rarely used together with other characters; there are some “Hub” characters, such as “Qi”, “Yin”, and “Yang”, connecting to a various number of characters in the sentences. We listed the basic network features of the 30 typical TCM books which are divided into 10 categories. It can be found that the power exponents of books are close to 1.0; the biggest one is 1.3246 of the Valuable Prescriptions for Emergency, and the smallest one is 0.9499 of the Medical Complete Book, Ancient and Modern. The node degree distributions of these ancient books followwhere k is the degree of the node, is the ratio of the number of nodes with a degree of k to the total number of nodes, and is the power exponent which floats above and below 1.

(a)

(b)

(c)

In addition, the average path lengths of these networks are around 3, in which the largest one is 3.819 and the smallest one is 2.727. The clustering coefficients are distributed between 0.1 and 0.3 (Table 2). Comparing these ancient books with random networks of the same scale, it is found that their average path lengths are smaller than random networks and the clustering coefficients are larger than random networks. It means that TCM language networks conform to the small-world pattern [7].

3.2. Topological Homogeneity of TCM Basic Concept Groups

To validate the power of complex network approach to differentiate the semantic groups of basic TCM concepts from the language network, we calculated the cosine similarity of each node vector of 16 basic TCM concept groups (Table 3). We assumed that the basic TCM concept groups, such as these concepts of five elements, would have similar values for the centrality measures, which would reflect their similar semantic importance in the language network from the topological measures. The results showed that most of the basic TCM concept groups in basic theoretical books (e.g., ICH) are more similar to each other than those of random controls (Table 4), which indicated that these basic TCM concept groups display a kind of linking homogeneity reflecting their close category semantic similarities. For example, the five elements concept category includes Mu, Huo, Tu, Jin, and Shui as closely related members. We found that, in the ICH book (Figure 2(c)), the cosine similarity of these five elements ranges from 0.0499 to 0.2786 with a rather high value of very narrow variance (mean: 0.1498 + std: 0.0669). This demonstrated the central role of the concepts in the five elements category for TCM and the categorical homogeneity of these five concepts. In addition, these category similar concepts could be identified by community detection methods due to their similar connection patterns in the context of network. For example, in ICH network, using FUA (see methods), we could identify the concept groups as same communities, such as “Yin and Yang”, “the five elements”, “the five notes”, and “the five colors” from the whole network (Figure 3).

However, the results were different for those clinical books (e.g., TCPD). The cosine similarity of the basic TCM concept groups did not tend to show homogeneous patterns. This might be due to the differently focused subjects of these books. For example, TCPD is mainly focusing on the manifestations of six types of syndromes and their regularities of herb treatment.

3.3. Diversity of TCM Language Networks

To further investigate the distinct topological patterns involved in different TCM language networks, we screened the links whose weights are >10 in ICH and TCPD and regarded the related nodes (Chinese characters) as key concepts in these two books (Figure 4). It is illustrated that the key concepts in ICH mainly include the basic theoretical characters in TCM, such as “Yin/Yang”, “the five elements” and its associated concepts, quantifiers, emotions, and pulse (Figure 4(b)). In contrast, although several basic theoretical concepts, such as “Yin/Yang”, are still included in TCPD as the key concepts, most of the others are related to herb prescriptions and symptoms (Figure 4(b)). These results indicated the distinct category of knowledge delivered in these two books. It is well known that ICH ensembles the basic theories of TCM, while TCPD is recognized as a representative clinical book focusing on disease manifestations, pathologies, and their corresponding herb prescriptions.

(a)

(b)

3.4. Exploring the Specific Semantic Intensions of Core TCM Concepts

To identify the specific meaning of a given concept, we would like to see what exactly words or phrases it occurred. The TCM language network could give help to this investigation. It is well known that some basic concepts, such as “Qi” and “Yin and Yang”, are of great significance to TCM; however, the connotations of these concepts are rather complicated [28–30]. We constructed an integrated language network with various character triples derived from 30 ancient TCM books, which finally consists of 6118 nodes and 381467 links. Here, we extracted the 1-order neighborhood of a given node and took four concepts, namely, “Qi”, “Yin/Yang”, “Xie”, and “Li” for demonstrations (Figure 5). It is interesting that for the basic concept of “Qi”, there are about 1864 characters (nodes) directly connecting to this node, in which the characters, such as “Yang”, “Xue”, “Yin”, “Yuan”, “Zheng”, and “Jing” together with “Qi” represent the main types of “Qi” recognized in TCM theories. The other connecting characters to “Qi” obtained the various manifestations and pathologies, such as “XiaQi”, “QiNi” and “QiXu”; “XieQi”, “HanQi”. Although these concepts are usually adopted by professional TCM practitioners, our network results could grasp and demonstrate the global connecting characters for TCM researchers. Similarly, we obtained 1736 characters related to “Yin and Yang”, which could form different “Yin and Yang”-related basic concepts, such as meridian-related concepts (e.g., “TaiYin”, “TaiYang”, “YangMai”, and “YinMai”), syndrome-related concepts (e.g., “YinXu” and “YangXu”). For the character “Xie”, our network approach clearly showed two types of semantics involved. One type of concept is related to different manifestations, such as “XieXie”, “TuXie”, “ShuiXie”, and “FengXie”. Another type of concept is related to the principles for prescriptions including “XieXin” and “XieHuo”. However, the concepts related to “Li” are only associated with disorders or diseases, such as “XueLi”, “NueLi”, “GanLi”, “LiChang”, and “LenLi” ¹. The rigorous evaluation of these related concepts would help with the precise understanding of the manifestations related to “Li” and improve distilling the high-value disease or prescription knowledge from TCM ancient literatures.

(a)

(b)

(c)

(d)

4. Discussion

Medical concepts constitute the basic knowledge framework of TCM theories devoted to the clinical observation of the complicated manifestations and their understanding of the underlying pathologies from TCM perspectives. Therefore, the development of TCM terminologies even with international translations is an important task in TCM field [31–33]. However, as most TCM concepts derived from ancient textbooks, it is difficult for contemporary practitioners to definitely grasp the whole meanings and connotations in the framework of TCM theories, in which the semantic diversity of a specific TCM concept is one of the key issues. Language network proposes an efficient approach to investigate the semantic properties of concepts of words in large-scale text corpora [34]. The application of complex network in linguistics has made it possible for us to adopt real network analysis tools in ancient TCM book studies. Unfortunately, the current researchers of TCM are mostly those with medical science background, who usually concern themselves with clinical medicine. It lacks some approaches, which focus on ancient TCM books’ concepts, are not only helpful to the research on basic theory of TCM but also helpful to nonprofessionals understanding the basic concepts.

Ancient TCM books are carriers of Chinese medicine knowledge and have great significance for the entire Chinese civilization [35]. In this paper, we analyzed TCM books in the form of network and explored some characteristics of ancient language networks. First of all, the node degree distributions, average path lengths, and clustering coefficients of the networks showed that TCM character-language networks follow a kind of scale-free and small-world networks. Secondly, we analyzed the basic concepts in ancient TCM book networks and found that these concepts play special roles in language networks. Furthermore, we extracted key TCM concepts of each book and found that the key concepts in different categories of ancient books have obvious differences. Finally, we drew a conclusion that Chinese medicine concepts such as “Qi” have rich medical connotations in ancient books.

There are several limitations in our manuscript. Firstly, we only constructed dozens of language networks, which might influence the extensions of the obtained results to more general context. In addition, the character-based 2-gram modeling also limits the investigation capability of the language network for semantic issues. Secondly, although most TCM basic concepts could be grasped by single character (e.g., Qi), there exist many key concepts, such as those of acupuncture points, herbs, and disorders, which would necessarily be represented by words or phrases to further explore their semantic regularities. Furthermore, it is notable that network approach is adept in investigating the global patterns of a given domain, which could be combined with other data analysis methods (e.g., association rules) to generate more specific results to deliver TCM meaningful knowledge.

5. Conclusion

In summary, we found that the degree distribution of ancient TCM book networks is consistent with power law distribution and small-world patterns. In addition, similar concepts in ancient books are displayed and linked closely. Moreover, we realized that there are essential differences in language and conceptual patterns between theoretical and clinical books. To sum up, the exploration of ancient TCM books provides an effective method to identify the underlying conceptual meanings of particular concepts conceived in TCM theories and clinical operations.

Data Availability

Data used in this paper are found at https://gitee.com/zouqunsheng/ancient-tcm-books.git.

Conflicts of Interest

The authors declare that they have no conflicts of interest regarding the publication of this paper.

Acknowledgments

This work was partially supported by the National Key Research and Development Program, China (2017YFC1703506, 2017YFC1703502, 2017YFC1703505, and 2020YFC0841600), the Special Programs of Traditional Chinese Medicine China (JDZX2015168, JDZX2015171, and JDZX2015170), and the Fundamental Research Funds for the Central Universities, China (no. 2018JBZ006).

References

P. Jiang, Y. Li, Y. Zhou et al., “Reflections on the collation of ancient books of traditional Chinese medicine,” Chinese Medicine Modern Distance Education of China, vol. 16, no. 6, pp. 62-63, 2018.
View at: Google Scholar
F. Meng, W. Shang, S. Li et al., “Classification systems of ancient books on traditional Chinese medicine and their evolution,” Chinese Journal of Medical Library and Information Science, vol. 24, no. 9, pp. 62–66, 2015.
View at: Google Scholar
T. Yu, J. Li, Q. Yu et al., “Knowledge graph for TCM health preservation: design, construction, and applications,” Artificial Intelligence in Medicine, vol. 77, pp. 48–52, 2017.
View at: Publisher Site | Google Scholar
J. Xie and C. Jia, “Types and functions of metaphorical language in internal classic of huang di,” Acta Chinese Medicine and Pharmacology, vol. 39, no. 1, pp. 1–4, 2011.
View at: Google Scholar
X. Zhou, Y. Peng, and B. Liu, “Text mining for traditional Chinese medical knowledge discovery: a survey,” Journal of Biomedical Informatics, vol. 43, no. 4, pp. 650–660, 2010.
View at: Publisher Site | Google Scholar
M. E. J. Newman, “The structure and function of complex networks,” SIAM Review, vol. 45, no. 2, pp. 167–256, 2003.
View at: Publisher Site | Google Scholar
S. H. Strogatz, “Exploring complex networks,” Nature, vol. 410, no. 6825, p. 268, 2001.
View at: Publisher Site | Google Scholar
H. Liu, “Linguistic Complex Networks: a new approach to language exploration,” Grundlagenstudien Aus Kybernetik Geisteswissenschaft (grkg/Humankybernetik), vol. 52, pp. 151–170, 2011.
View at: Google Scholar
S. Li, B. Zhang, D. Jiang et al., “Herb network construction and co-module analysis for uncovering the combination rule of traditional Chinese herbal formulae,” BMC Bioinformatics, vol. 11, no. 11, p. S6, 2010.
View at: Publisher Site | Google Scholar
C. van Borkulo, L. Boschloo, D. Borsboom, B. W. J. H. Penninx, L. J. Waldorp, and R. A. Schoevers, “Association of symptom network structure with the course of depression,” JAMA Psychiatry, vol. 72, no. 12, pp. 1219–1226, 2015.
View at: Publisher Site | Google Scholar
R. Goekoop and J. G. Goekoop, “A network view on psychiatric disorders: network clusters of symptoms as elementary syndromes of psychopathology,” PLoS One, vol. 9, no. 11, Article ID e112734, 2014.
View at: Publisher Site | Google Scholar
Y. Bai, J. Wang, J.-P Wu et al., “Review of evidence suggesting that the fascia network could be the anatomical basis for acupoints and meridians in the human body,” Evidence-based Complementary and Alternative Medicine, p. 2011, 2011.
View at: Google Scholar
H. Wan, M.-F. Moens, W. Luyten et al., “Extracting relations from traditional Chinese medicine literature via heterogeneous entity networks,” Journal of the American Medical Informatics Association, vol. 23, no. 2, p. 356, 2016.
View at: Publisher Site | Google Scholar
P. F. Brown, P. V. Desouza, R. L. Mercer et al., “Class-based n-gram models of natural language,” Computational Linguistics, vol. 18, no. 4, pp. 467–479, 1992.
View at: Google Scholar
W. B. Cavnar and J. M. Trenkle, “N-gram-based text categorization,” 1994.
View at: Google Scholar
J. H. Martin and D. Jurafsky, “Speech and language processing: an introduction to natural language processing, computational linguistics, and speech recognition,” Pearson/Prentice Hall, vol. 23, 2009.
View at: Google Scholar
M. Brede, “Networks—an introduction,” Artificial Life, vol. 18, no. 2, pp. 241-242, 2010.
View at: Google Scholar
M. Newman, “Power laws, Pareto distributions and Zipf's law,” Contemporary Physics, vol. 46, no. 5, pp. 323–351, 2005.
View at: Publisher Site | Google Scholar
U. Brandes, “A faster algorithm for betweenness centrality,” The Journal of Mathematical Sociology, vol. 25, no. 2, pp. 163–177, 2001.
View at: Publisher Site | Google Scholar
D. J. Watts and S. H. Strogatz, “Collective dynamics of “small-world” networks,” Nature, vol. 393, no. 6684, pp. 440–442, 1998.
View at: Publisher Site | Google Scholar
Q. Jane, “Traditional medicine: a culture in the balance,” Nature, vol. 448, no. 7150, p. 126, 2007.
View at: Google Scholar
A. Grover and J. Leskovec, “node2vec: scalable feature learning for networks,” Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, vol. 4, pp. 855–864, 2016.
View at: Google Scholar
J. F. Box, “Guinness, Gosset, Fisher, and small samples,” Statistical Science, vol. 2, no. 1, pp. 45–52, 1987.
View at: Publisher Site | Google Scholar
J. L. Devore and R. Peck, “Statistics: the exploration and analysis of data,” Duxbury Press, vol. 4, 2005.
View at: Google Scholar
V. D. Blondel, J. L. Guillaume, R. Lambiotte et al., “Fast unfolding of communities in large networks,” Journal of Statistical Mechanics, vol. 2008, no. 10, pp. 155–168, 2008.
View at: Publisher Site | Google Scholar
C. M. Bernat and R. V. Solé, “Universality of Zipf’s law,” Journal of Statistical Mechanics, vol. 82, no. 1, Article ID 011102, 2010.
View at: Google Scholar
B. Albert-László, “Scale-free networks: a decade and beyond,” Science, vol. 325, no. 5939, pp. 412-413, 2009.
View at: Google Scholar
X. Dong, “The meaning evolution of “material force” and the theory in traditional Chinese medicine,” Journal of Liaoning Medical College: Social Science Edition, vol. 14, no. 4, pp. 63–66, 2016.
View at: Google Scholar
M. Yan and Y. Wang, “Objectization of conceptual and state of affairs of spatiotemporal sequence by qi’s reference in traditional Chinese medicine,” Journal of Beijing University of Traditional Chinese Medicine, vol. 41, no. 9, pp. 717–725, 2018.
View at: Google Scholar
Z. Wang, “Study on the essence of Yin and Yang,” Traditional Chinese Medicine and Related Issues, vol. 41, 2014.
View at: Google Scholar
Z. Li, “Standardizing English translation of traditional Chinese medical terminology:an analysis of the Concepts,Principles and methods concerned,” Chinese Translators Journal, vol. 29, no. 4, pp. 63–70, 2008.
View at: Google Scholar
J. Zhu, “Standardization of terms of traditional Chinese medicine (TCM) and modernization and internationalization of TCM,” China Journal of Traditional Chinese Medicine and Pharmacy, vol. 1, pp. 6–8, 2006.
View at: Google Scholar
Z. Xie, “Discussion on English translation of traditional Chinese medicine terms,” Chinese Journal of Integrated Traditional and Western Medicine, vol. 9, pp. 706–709, 2000.
View at: Google Scholar
C. Biemann, S. Roos, and K. Weihe, “Quantifying semantics using complex network analysis,” Proceedings of COLING, vol. 2012, pp. 263–278, 2012.
View at: Google Scholar
Y. Fu, G. Liu, B. Li et al., “Digital research on ancient books of traditional Chinese medicine,” Chinese Journal of Information on Traditional Chinese Medicine, vol. 6, pp. 563-564, 2004.
View at: Google Scholar

Copyright

Copyright © 2020 Qunsheng Zou et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

PDF Download Citation

Download other formats

Order printed copies