1. Introduction
This paper explores TV webpages’ network analyses tocompare and to figure out at the level of national TV station & rsquo;s hyperlinks structures. In the big data period, the program viewership through web has leaped up numbers. In the age of cyberspace, not only description on identity that comes in real spaces, but also through data from a typical cyber-world is very important. Moreover, it has beenimportant research topics to find out the characteristics ofindividual TV networks through network analysis. In the big-data-hyperlink structure, the research would figure out the TV-web hyperlink structure transformation from single mode to multi-modes.
“Webmetrics” are newly appeared in the mid/late 1990s in the 20th century to conduct researches. The researches areapplied to Web related field, but the conclusion could bedrawn due to effects of multilateral elements such as countries, regions, languages, economy, and culture. Thereafter, as studies developed gradually, distinctive characteristics of the Internet were found so that it could be seen that studies should be conducted with a combination of the literature research method and diverse scientific study methods.
Among the foregoing study methods, the Internethyperlink analysis method received great attention. Through scholars & rsquo; researches on web-metrics, which are one of distinctive characteristics of the Internet, it could be seen thathyperlinks between web and web-networks could re inforcethe exchanges and relationships between individuals and organizations, and between or institutes that act on the cyberspace. Moreover, researchers focused on hot topics that attract all people’s attention for a certain period of timecould be identified so that even core websites could be foundout. In addition, through the results of empirical studies, it shows that, the state of the web itself is not confusing unlike expectation but has structures that are stable and good inquite a large range according to simple mathematical rules. Broder et al(2000) published the results of verification of approximately 200 million websites and 1.5 billion hyperlinks were considered interactive, at least 90% of sample websites consisted of single hyperlinks. The probability for a hyperlink to exist between randomlyselected two websites was shown to be as high as 24% and any two websites were connected with each other 16 hyperlinks on average[1]. Therefore, the method of studying hyperlink structures is essential and reliable.
As terrestrial TV networks have been providing diverse information and services. Therefore, the audience couldeasily access information and program services, they have become people’s teachers and friends while becoming animportant part that cannot be replaced in people’s daily lifein the process of giving knowledge and pleasure. In thissense, studies on terrestrial TV networks can be regarded to be highly valuable. When the structures and characteristics of the TV networks’ home pages are known, better ones among the network structures can be generally determined and the contents added or deleted can be seen from the viewpoint of producers of TV networks’ home pages so that the direction of improvement can be presented.
In addition, although some studies already analyzed hyperlinked websites of domestic terrestrial TV networks in South Korea, there are not much in the home pages ofterrestrial TV networks. Since different countries are quitedifferent in many fields such as economy, culture, and language, the network structures and types of websites hyperlinked to the home pages of those terrestrial TV networks were expected to show large differences. Ratherthan comparative studies of the home pages of domesticterrestrial TV networks that have been operated in similar Internet and press environments in South Korea, comparativestudies of international terrestrial TV networks may stimulate the birth of more novel and unexpected good ideas.
The present study try to visualize the structures of websites hyperlinked to the official home pages of TV networks to analyze the hyperlink network structures ofindividual TV networks to examine the connection relationships between the TV networks. In addition, the types of those websites that are hyperlinked to the official homepages of individual TV networks will be identified to examine if there are differences between them through content analyses. Based on the results of the study conducted as such, significant opinions for the direction of improvement of the home pages of international terrestrial TV networks will be presented to help those international terrestrial TV networks find ways to survive in the Internet environments where they compete fiercely with each other.
2. Theoretical discussions
2.1 Webometrical Analysis
Scholars began to study the phenomenon of interactions between Web structures through citation networks and hyperlinks from the middle of the 1990s and gave the name & lsquo; Webometrics & rsquo; to this new field. According to Bjὄrneborn(2004), ‘Webometrics’ is an academic field to study the construction and use of web information resources and the quantitative aspect of web structures and technologies using metrical bibliographic and informetrical approaches[2]. According to this definition, major study areas of ‘Webometrics’ can be largely divided into four categories; 1) studies related to the analysis of the contents of web pages, 2) studies related to the analysis of web link structures, 3) studies related to web usability analysis through the analysis of users’ log files, and 4) studies related to web technology analysis for the performance of search engines.
In Webometrics, the connection between one website toother web papage are divided into ‘in-links’ and ‘ out-links. ’ The relationships between webs connected with each otherappear are based on hyperlinks. The link relationships between ‘nodes’ that mean analysis units are shown in 1> that shows the number of connections in a five-nodenetwork-based information content. The same as in 1>, the network is getting more pruned from the simplelinearity, the information content based on the number of connections increases(Bonchev & Buck(2005). The complexity of a well-defined network can be used to quantify the structural characteristics of a particular complexity. Through this anlysis, this study not only helps to understand the network between web sites through the hyperlink of the broadcasting company 's homepage, but also the value and meaning of the terrestrial broadcaster' s homepage as big data, meaning of hyperlink structure I intend to draw implications.
(Figure 1) Link Structure and Vertex Degree in Number of connections in a five-nodenetwork-based information content
2.2 Network Analysis Technique
Network analysis techniques are methods of analyzing the social relationships between nodes such as centrality ordensity using the connection relationships betweenindependent nodes existing on networks. The applied network analysis methods include ‘degree centrality,’ ‘betweennesscentrality, ’ ‘closeness centrality,’ and ‘eigenvector centrality.’
2.2.1 Degree Centrality
Degree centrality is an indicator. It is to measure the importance of each node in a network by measuring the number of nodes to which the node is directly connected. Values calculated as degree centrality indicate the number of nodes to which each user is connected on the network.
2.2.2 Betweenness Centrality
Betweenness centrality is an indicator used to measurehow much the relevant node plays the role of anintermediary. A high index of betweenness centrality affects greatly information flows in the network[3]. Therefore, othernodes become to be much interested in the information of this node in many cases.
2.2.3 Closeness Centrality
Closeness centrality is an indicator used to measure the centrality considering not only direct connections but alsoindirect connections because the effects of nodes on thenetwork cannot be easily identified. In this method, distancesto all nodes connected with the node being measured arecalculated to the shortest distance. When closeness centrality is higher, information power, influence, and position in thenetwork can be more easily secured and accessed.
2.2.4 Eigenvector centrality
Eigenvector centrality is an indicator used to find the most influential node in the network. It is affects anothernode and the affected node affects many other nodes, the first node in such a node circulation can be said to be highly influential and ‘eigenvector centrality’ is used to find the first node.
2.3 Classification of website types
Currently, numerous types of website services exist and they are unceasingly differentiated. However, noclassification system that organized such types of Internetservices based on empirical data exist thus far and classification of websites is being attempted using diversecriteria such as profit creation methods.
Trochim (1996) classified five types of websites; 1) information provision, 2) education & training, 3) commerce & amp; advertisement, 4) entertainment, and 5) communication. Alexander and Tate (1999) classified as six types; advocacy, business, information, news, individuals, and entertainment. Kim Gyeong-Ja (2002) divided websites into four types; portal, community, commercial transactions, and information provisions to find and analyze customer service dimensions.
In a study conducted by Kang Eun-Jeong(2003), unlikemost cases of website classification where websites were classified according to the purposes of websites, websites were classified using major contents into six types; Identity, Information, Shopping, Learning, Community, and Entertainment. Tarafdar and Zhang(2006) classified websites based on characteristics, functions, and evaluation into five types; portal & search engines, retails, entertainment, news & information, and financial services.
3. Research Methods
3.1 Study Subjects
As Internet technology and communication technology have been rapidly developing, traditional media and new social media exchange information with media content consumers through the Internet. To understand howinformation flows in social network spaces, studying the information flows through the hyperlink analysis method is badly necessary.
TV networks should continuously receive in-links from numerous websites to provide services including news, business, and entertainment to general people, Since TV networks generally have long histories and extremely largenumbers of viewer, they are considered to have very largesocial influence[4]. The characteristics of international TV networks were expected to be different to some extent because they are affected by many different elements such aspolitics, economy, culture, and language that are different by country [5].
In the present study, to examine the characteristics of international TV networks and associations and differencesamong them, four TV networks; ABC, BBC, NHK, and KBS that have relatively high awareness and influence among Koreans were selected to conduct hyperlink analyses. Concrete research problems are as follows.
Research problem 1: How are the hyperlink structures of ABC, BBC, NHK, and KBS in the homepage structureconstructed?
Research problem 2: Are the types of websites hyperlinked to the official home pages of TV networks ABC, BBC, NHK, and KBS different from each other?
3.2 Study process
In the present study, two different study methods were used to solve the two research problems.
3.2.1 Hyperlink data collection program-VOSON
VOSON is a hyperlink network data collection tool provided by Australian Demographic and Social Research Institute and Australian National University. Currently, VOSON has been installed on diverse network analysis programs to provide data for hyperlink network analysis.
In the present study, in order to completely figure out the hyperlink structures of the official home pages of international public TV networks, all values were set to the maximum as shown in the “Fig. 2”.
(Figure 2) Voson Data Collection
The URLs of the official home pages of individual TV networks are follows.
The URL of ABC is 'www.abcnews.go.com', that of BBC is 'www.bbc.com', that of NHK is 'www.nhk.or.jp', and that of KBS is 'www.kbs.co.kr '. One things that is noteworthy here is that when we entered 'abc' in Naver, the official website with the URL address 'www.abc.go.com' was shown but this website showed only those contents that are related to ABC TV network’s TV programs and was completely different from the official home pages of other public TV networks (BBC, NHK, KBS) that provide diverse servicessuch as news and bussiness. Therefore, websites weresearched again to find the official home page of ABC TV network of which the URL address is 'www.abcnews.go.com' was found and this website was selected as a study subject because this website satisfied the criteria for TV networks mentioned earlier.
(Table 1) Home Page Picture of the Four TV Networks
3.2.2 Network analysis program-NodeXL
NodeXL is a network analysis tool developed by Microsoft that is operated based on Excel to supportinformation extraction from websites that provide SNS services, the use of e-mail Outlook, and personal network analysis [6]. Since this is based on Excel, it can be used fordata used in other network analysis tools and can expressanalysis results in diverse diagrams made utilizing Excelfunctions in addition to the Node-Link form(Bonsignore et al, 2009). The NodeXL Tool enables adding network graphs in the form of charts to
Excel spread sheets that are used everywhere and also enables both beginners and experts to easily conduct network analysis (Smith et al, 2009)[7].
In the present study, data in the form of Excel charts were received through the data collection tool VOSON installed on the NodeXL program and were visualized in NodeXL to obtain the figures of the network structures of websites hyperlinked to the official home pages of four international TV networks; ABC, BBC, NHK, and KBS.
This paper applied the module of centrality typesincluding degree centrality, closeness centrality, betweencentrality, and eingenvector centrality.
(Table 2) Formula of Degree Index
In addition this paper classified the type of industryclassification to classified the business type centralities including portal, information, business, entertainment, personal, advocation, and communication.
3.3 Classification of the types of hyperlinked websites
In the present study, referring to the previous studies regarding website type classification presented in theoretical background, websites hyperlinked to the official home pages of four international TV networks; ABC, BBC, NHK, and KBS were classified into a total of eight types; portal and search engine, business, personal, news and information, entertainment, advocacy, communication, and unidentifiable [13][14][15][16][17][18][19][20].
To review concretely, first, those websites that provide wide-ranged information on diverse subjects but link and connect to other websites when titles regarding information have been clicked were classified into the 'portal and searchengine' type, second, those websites that had purposes suchas retail, shopping, retail, shopping, commercial transactions, finance, advertisement, or service sales promotion were classified into the 'business' type, third, those websites thathad personal purposes such as personal dilettante lives, emotion expression, diversion, etc. were classified into the' personal ' type, fourth, those websites that provided diverserecent information such as news, education, learning, and weather produced by themselves and did not receive in-links from other websites were classified into the 'news andinformation ' type, fifth, those websites that provided information or content centering on music, dramas, movies, entertainment, and games were classified into the' entertainment ' type, sixth, websites of non- profitorganizations, administrative organizations, and NGOs were classified into the 'advocacy' type, seventh, those websites for the purpose of exchanges such as SNS, cafe, bulletinboards, and e-mails were classified into the ' communication ' type, and finally, those websites that could not be opened due to changes in URL addresses or the expiry of the serviceperiods were classified into the 'unidentifiable' type [21][22][23][24][25][26][27].
In the present study, hyperlinked websites were coded one by one based on the classification criteria for the eightwebsite types after entering them to see their contents and the contents were analyzed with the data from them. Theresults of chi-square tests and correspondence analyses wereanalyzed to find differences in the types of websites hyperlinked to the official home pages of four TV networks; ABC, BBC, NHK, and KBS[28][29][30][31][32][33].
4. Study Results
4.1 Results of analysis of ABC, BBC, NHK, and KBS in the hyperlink structures of the official home pages of TV networks
‘Research problem 1’ is exploring the hyperlink structures of the official home pages of four TV networks. To this end, a total of 4378 collected nodes were analyzed using a NodeXL processor and the connection structures among the hyperlinked-websites were visualized.
In the analysis results, a total of 4,649 links were foundamong a total of 4,378 nodes and the density of the graphwas 0.00024. The average total connection degree (sum ofin-degree and out-degree) was 2, the average life in-link connection degree (in-degree) was 1, the average out-link connection degree (out-degree) was 1, the average betweenness centrality value was 9533.396, the average closeness centrality value was 0.000, and the average eigenvector centrality value was 0.000.
In the results of calculation of the values of the degrees of connections to each website conducted for connectiondegree centrality analysis, the total number of links of ABC was 1407, it is the largest. The second is NHK at 1231, BBC at 1119, and KBS at 894.
BBC, NHK, and KBS were overwhelmingly larger thanthose of other websites but also the numbers of in-links werelarger than the numbers of out-links, it could be seen that all the four TV networks had large influences in the hyperlinked network and were playing central roles.
(Table 3) Differences in the types of the official home pages of the four TV networks
(Figure 3) Network maps of websites hyperlinked to the official home pages of the four TV networks
In addition, new websites that must be watched for connection degree centrality were found. Three websites;'.angelfire.com/' , 'ria.ru/', and '.geocities.co.jp/' had the largest numbers of links except for the four TV networks, ABC, BBC, NHK, and KBS because the sum of the numbers of in-links and out-links of all of these three websites werecalculated to be 4. There are visualized figure all the links of three websites were out-links and were found to have beenhyperlinked to ABC, BBC, NHK, and KBS.
The values of betweenness centrality were calculated tomeasure the degree to which a website is located betweenother websites in the network. To review the results, among the four TV networks, the level of the betweenness centrality of ABC (9804143.880) was the highest followed by NHK(8524331.590), BBC (7845225.120), and KBS (6640777.410). In addition, the three websites mentioned above(' http://www.angelfire.com/' , 'http://ria.ru/', and ' http://www.geocities. co.jp/') recorded the highest betweenness centrality value (302235.152) next to ABC, NHK, BBC, and KBSindicating that they were at important positions to arbitrate or mediate other websites in the entire hyperlink network.
The closeness centrality values is connected to all otherwebsites in the hyperlink network. Although all of the values were indicated as 0.000 in the table presented above because the closeness centrality values were set down to three places of decimals, actual values obtained were 'http://www.angelfire. com/'(0.000114), 'http://ria.ru/'(0.000114), 'http://www.geocities. co.jp/'(0.000114), ABC(0.000108), and BBC(0.000105) indicating that these websites were closest to other websites.
The eigenvector centrality values were the most influentialin the entire hyperlink networks of the official home pages of the TV networks. In the results, the ABC’s eigenvectorcentrality value (0.010) was shown to be the highestindicating that the official home page of ABC was the most central website in the hyperlink network. The BBC’seigenvector centrality value is 0.005, and the NHK’seigenvector centrality value (0.004) were produced similarly indicating that the official home pages of these two TV networks were relatively more central than other websitesexcept for the official home page of ABC. In addition, the 3 website that continuously showed high centrality valuesearlier showed the same eigenvector centrality value (0.001) as KBS this time indicating that they were relativelyperipheral websites of the hyperlink network.
The result shoes that the official home page of ABC was judged to be the most influential website and it was found that the three newly appeared websites were at important positions and that all of them were hyperlinked to ABC, BBC, NHK, and KBS as shown in the Figure that visualized the hyperlinks. Also the results confirmed that none of the websites were large in scale. The reason why theynevertheless occupied important positions is that they werelinked to all the four very highly influential TV networks [14].
Therefore, the strategy to link with more influential websites is considered more appropriate for the expansion of the influence of a website rather than linking with many websites.
4.2 Results of analysis of differences in TV networks 4.2 ABC, BBC, NHK, and KBS in types among the websites hyperlinked to the official home pages.
The study confirmed that chi-square tests were conducted to examine the differences in types among the websites hyperlinked to the official home pages of international TV networks ABC, BBC, NHK, and KBS. The results of the chi-square analyses indicated that there were differences in types among the websites hyperlinked to the official homepages of different international public TV networks (χ 2=236.481, df=18, p=0.000).
The results of the relationship between the variable ‘ TVnetworks & rsquo; containing the four TV networks; ABC, BBC, NHK, and KBS as lower categories and the variable ‘ website types & rsquo; containing the seven website types; portal and searchengine, business, personal, news and information, entertainment, advocacy, and communication as lowercategories are as follows.
According to the results of examination of information on the repeated calculations in the correspondence analysis, the average eigenvalue finally fitted through a total of 51 repeated calculations is 1.139. In addition, 60.9% of the variance of the data was explained by one dimension and 53.0% was explained by two-dimension. Therefore, 57.0% of the variance was explained by the two different
The combination system of the coordinates of the categories under 'TV networks' and 'website types' forindividual dimensions are schematized as follows. In &ld quo;Fig. 4&rd quo;, all the categories under 'TV networks' are gathered in the right top area in the case of one dimension and the categories under 'website types'; advocacy, portal and searchengine, news and information, business, personal, entertainment, and communication are distributed in the case of two dimension. On reviewing the figure concretely, it could be seen that, NHK were closely related with portal & search engines and business, KBS showed the shortest distanced to business and personal, ABC was the closest tonews and information, and BBC showed short distances to news & information and personal. On the other hand, the categories: advocacy, entertainment, and communication under website types showed long distanced to all the four TV networks.
(Figure 4) Combination chart of category points under 'TV networks' and 'websites types'
(Table 5) Coordinates of categories under ‘ TVnetworks & rsquo; and ‘websites types’
5. Conclusion and Implication
5.1 Study conclusion
In the present study, based on the numerous network analysis methodologies presented in Webometrics studies, thenetwork analysis program NodeXL and the content analysis program SPSS were combined to figure out the structures and types of those websites that were hyperlinked to the official home pages of the four international TV networks, ABC, BBC, NHK, and KBS and simple comparative studies were conducted. Through the studies, differences in the characteristics and types of hyperlink connection structuresamong the home pages of the individual TV networks werefound and concrete conclusions were drawn as follows.
First, to review the numbers of nodes that mean the numbers of websites associated with the four TV networks, the numbers of nodes associated with ABC, BBC, NHK, and KBS were1,377; 1,095; 1,216; and 894 respectively. Therefore, the ABC can be said to be the most influential followed by NHK, BBC, and KBS. KBS had the smallest number of nodes and was a little less influential compared to the other three TV networks. In fact, ABC and BBC areless restricted by language because all the contents of the websites are in English and English has been continuously accepted as an international current language throughout the world and has been designated as a mother tongue in many regions. In addition, since the USA and the UK have beenhighly influential internationally for long and their TV networks had been relatively highly influential[34][35][36 ][37][38]. Given these conditions, the abovementioned results are considered natural. The weak influence of KBS has beenexpected because the entire content of its home page is in Korean, Korean is used only in Korea, and the area of Korea is not large. However, unexpectedly, NHK TV network in Japan having similar objective conditions to those of Koreashowed a larger number of associated websites clearly indicating that NHK had good network structure at a quite high level[39][40][41][42][43].
Second, the result shows that the number of out-links of ABC was the largest at 30 and that of KBS was at 0. Whenthe websites out-linked to ABC were accessed and reviewed, it could be seen that many external websites independently established by ABC were found[44][45][46][47][48]. That is, rather than operating a single website, ABC intentionallymade diverse and subdivided websites of it and was reinforcing interactions between the websites with a view to improving the influence of the entire ABC TV network. Infact, the largest influence of ABC TV network because such actions could exert positive effects to some extent. In contrast, KBS had no out-link and this could be regarded to be one of the reasons for the lowest influence indicating the necessity for KBS to learn many things from ABC[49][50][51][52][53].
Third, on reviewing the network structures indicated in &ld quo;Fig. 3”, many new facts could be seen. First, the number of links between ABC and BBC was the largest and the number of links between BBC and NHK was shown to besimilarly large. To review concretely, ABC was the most closely associated with BBC and much less
closely associated with NHK and KBS. BBC was closely associated with both ABC and NHK but was hardly associated with KBS. NHK showed high levels of association with BBC and KBS and a low level of association with ABC. KBS was low level associated with ABC or BBC, but was closely associated with NHK. The close association between ABC and BBC can be easily understood because they are geographically close and have similar socialenvironments. Of course, KBS and NHK were also closely associated for similar reasons. However, it should be noted that both BBC and NHK were closely associated with two TV networks despite long geographical distances and largecultural differences so that they can be regarded to be progressive and open TV networks. Fourth, the study tested the types among the official home pages of the four TV networks were examined. The results of chi-square tests showed the order of the numbers of the types of all websites hyperlinked to the TV networks from the type with the largest number of websites as 'news and information' >' portal and search engine' > 'business' > 'personal' >' communication ' > 'entertainment' > 'advocacy'. Among the types, the type 'news and information' had anoverwhelmingly large number of websites and the highestratio. In the category point combination chart drawn through correspondence analysis, ABC showed the longest distance to the 'entertainment' type because it mainly used the homepage for 'news and information.' BBC showed somewhat longer distances than the reaming three TV networks and was farthest from the 'entertainment' type because it mainly used the home page for 'news and information' and ‘personal & rsquo; purposes. NHK had the largest number of types of websites surrounding it and was farthest to ‘entertainment & rsquo;and 'communication' types because it mainly used the homepage for 'portal and search engines' and 'business.' KBS was farthest from the 'entertainment' type because it mainly used the home page for 'business' but its distance to the' entertainment ' type was shortest among the four TV networks. When seen comprehensively, the types of websites hyperlinked to the official home page of NHK were relatively the most diverse.
Fifth, when the category point combination chart drawn through correspondence analyses was analyzed again afterchanging the viewpoint to see the results from the viewpoint of hyperlinked website types instead of TV networks, the' news and information' type was found to be closest to ABCTV networks, the 'portal and search engine', 'business', and ' personal ' types were found to be closest to NHK TV networks, the 'entertainment' and 'advocacy' types werefound to be closest to KBS, and the 'communication' type was found to be closest to BBC TV networks. The shortest distance means that the network has a larger number of that type of websites than other TV networks. This information may enable Internet users to efficiently select TV networkhome pages according to the types of information that they want to find out.
The sum of the results, the NHK TV network in web pages attracted the greatest attention. NHK TV networkcan be said to have constructed relatively the most rational and successful website networks among the TV networks. Itovercame it’s drawbacks to have more associated nodes than BBC while having 15 out-links, and the most diverse hyperlinked website types. The home page of KBS TV network that has been operated in similar environments to those of NHK showed relatively many shortcomings.
The above conclusions clearly should have some parts not consistent with facts because the analyses were conducted with data for a (or a certain) period of time becausehyperlinks are so dynamic and change frequently following changes in external environments so that conclusions cannot be easily drawn from the data collected at a time. If datacollected in many lots over a quite long period of time areanalyzed, more accurate results should be produced. Although data from different time zones could not be collected in the present study, the present study is considered meaningful in that it presented a new method of studying the structures and characteristics of TV networks’ home pages and in particular, it is considered more meaningful in that it presented a new direction of thinking about the way how to draw conclusions with analysis results.
5.2 Discussion and implications
The present study shows significant results, on the other hand it also has several limitations, which are examined concretely as follows:
First, when hyperlink data with VOSON search engine, allof the numbers of collection variables were set to the maximum to collect as many as possible websites hyperlinked to the official home pages of TV networks so that both directly connected websites and indirectly connected websites could be studied. Consequently, despite that only four TV networks were the data obtained exceeded 4,000.
Of course, the results should be general and highly reliable because large numbers of data are analyzed but the number of TV networks that can be analyzed should decrease. In the process of collection for the present study, data on a total of five TV networks including CCTV were intended to be collected at first but an e-mail was receivedindicating that the volume of data was too big to process through VOSON. Therefore, CCTV was deleted and the data of the remaining four TV networks were collected first and data on CCTV were collected separately. Thereafter, the two collected data files were combined into one file and attempts were made to analyze the file through a NodeXL processorbut the program was continuously shut down automatically because the data volume was too big. Eventually, only the data on four TV networks; ABC, BBC, NHK, and KBS wereanalyzed. Therefore, if many TV networks should be studied later, appropriate numbers of data should be set when the data are collected.
Second, as mentioned earlier, although the home page of ABC TV network is known to be 'www.abc.go.com', it was found that this website showed only those contents that are related to ABC TV network’s TV programs and was completely different from the official home pages of otherpublic TV networks (BBC, NHK, KBS) that provide diverseservices such as news and business when it was accessed. Therefore, the official home page of ABC TV network of which the URL address is 'www.abcnews.go.com' that had a similar structure to the home pages of three TV networks; BBC, NHK, and KBS was selected as a study subject. Although the foregoing decision unavoidably affected the study result to some extent, it is assumed to have been a relatively good decision in the unavoidable situation. Therefore, if ambiguous home pages appear later when study subjects are selected, those home pages should be carefully classified to determine study subjects with care based on the purpose of the study.
Third, longitudinal studies should be conducted on hyperlink networks. Many websites were unidentifiable because URL addresses of some websites were changed ordisappeared over time. In addition, in the present study, the data were collected on December 12, 2015 and demand fornews in the USA seemed to be higher than ever before because the presidential election was approaching affecting the study results to some extent. Therefore, long-term studies should be conducted to examine diverse situations of thenetworks hyperlinked to the home pages of individual TV networks and the processes of their changes over time. In addition, since the number of visitors to KBS may rapidly increase in a short time when a social event has occurred ata certain time, for instance, when the Sewol ferry disaster has occurred, such situations should be also considered when analyzing data.
If follow-up studies would supplement the threelimitations presented above and conduct more diverseanalyses, those studies should be more objective.
참고문헌
- A. Broder, R. Kumar, F. Maghoul, P. Raghavan, S. Rajagopalan, R. Stata, A. Tomkins, J. Wiener, "Graph structure in web", In Computer Networks, vol.33, pp.309-320,Jun.2000.https://www.cis.upenn.edu/-mkearns/teaching/NetworkedLife/broder.pdf https://doi.org/10.1016/S1389-1286(00)00083-9
- Bjorneborn, L. "Small-world link structures across an academic Web space: A library and information science approach", Doctoral thesis of Graduate School of Library and Information Science, Univeristy of Copenhagen, 2004. https://informationstudies.ku.dk/publications/list/?pure=en%2Fpublications%2Fsmallworld-link-structures-across-an-academic-web-space(9e644a39-6eef-4eef-b08b-ce8d6b9a084d).html
- Pei Yun-feng. "Contents recommendation method using betweenness centrality analysis", Master's thesis of the Graduate School of Computer And Information Science, KoreaUniversity,2011.http://www.riss.kr/link?id=T12707259
- Choi Chang-Hyeon. "A study on organizations' informal connection networks-utilization of analysis of social connection networks", Studies of Korean Society and Administration, vol.17, no.1, pp.1-23, 2006. http://www.kossrec.org/?page_id=474
- Hansen D.L., Shneiderman B., & Smith A.M. "Analyzing Social Media Networks with NodeXL", Amsterdam: Elsevier Inc. 2011.
- Hwang Gyeong-Ho. "Analysis of the hyperlink structures of TV networks' home pages: centering on comparative analysis of terrestrial and comprehensive programming channels", Doctoral thesis of the Graduate School of Press Information, SungKyunKwan University, 2015. http://www.riss.kr/search/detail/DetailView.do?p_mat_type=be54d9b8bc7cdb09&control_no=ef1511f6477d1be6f fe0bdc3ef48d419
- Jeong Yeong-Mi. "Analysis of the characteristics and structures of academic web spaces for activation of national information distribution", Seoul: Korea Institute of Science and Technology Information, 2007. https://www.kisti.re.kr/
- Kim Kam-Min. "web2.0 and social network", Institute for Information Technology Advancement academic information, 2005.http://www.iita.re.kr/main.it
- Kim Yong-Hak, Park Se-Wung, Jeon So-Yeong. "Analysis of the structure of online social connection networks", Informatization Policy, vol. 13, no. 4, 2006. https://www.nia.or.kr/
- Larsen, K. "Knowledge network hubs and measures of research impact, science structure, and publication output in nanonstructured solar cell research", Scientometrics, Vol.74, No.1, pp.123-142, 2008. https://link.springer.com/article/10.1007/s11192-008-0107-2
- Lee Chang-Gyun, Seong Min-Jun, Lee Yun-Bae. "Discovery of e-commerce enterprises, customer service cool trends: utilizing social network analysis NodeXL", Information Systems Review. Vol. 13. No. 1, 2011. http://acoms.kisti.re.kr/journal.do?method=journalintro&journalSeq=J000154
- Lee Jae-Yun. "Analysis of domestic academic field networks based on the present situation of journals for researchers' contribution", Journal of the Korean Society for information Management, vol. 25, no. 4. 2008. https://kosim.jams.or.kr/co/main/jmMain.kci
- Lee Jae-Yun. "Comparative study of weighted network centrality indexes of tnet and WNET", Journal of the Korean Society for information Management Vol. 30, No. 4, pp.241-264, 2013. https://kosim.jams.or.kr/co/main/jmMain.kci https://doi.org/10.3743/KOSIM.2013.30.4.241
- Lee Ju-Ho. "A study on weighted network centrality using the mutual citation index of statistics related journals", Master's thesis of the Graduate School of Statistics, Dongguk University, 2012.
- Otte,E., & Rousseau, R. "Social Network analysis: A powerful strategy, also for the information sciences", Journal of Information Science. Vol. 28, No. 6, pp.441-454, 2002. https://journals.sagepub.com/home/jis https://doi.org/10.1177/016555150202800601
- Park Chi-Seong. "Social network methodologies and their implications in administrative/policy studies", Chung-Ang University Spring Conference, 2010.
- Park H. W., Barnett G. A., & Nam I. Y. "Interorganizational Hyperlink Networks among Websites in South Korea", Networks and Communication Studies, Vol. 16, No.3-4, pp.155-174, 2002. https://www.researchgate.net/publication/228872128_Int erorganizational_hyperlink_networks_among_Websites_i n_South_Korea
- Seo Jeong-Ju, Kim Jin-Il, Kim Eun-Sang, Kim Yeong-Ho, Jeoformation Scientists and Engineers : Reang Ha-Wung, Kim Seong-Ryeol, Park Geun-Su. "Link structure Analysis of Korean web graph", Journal of Korean Institute of Inlity of computing and letter, Vol. 19, No. 1, pp.7-14, 2013.
- Bonchev, D., & Trinajstic, N. "Information theory, distance matrix, and molecular branching". The Journal of Chemical Physics, Vol. 67, No. 10, pp.4517-4533, 1977. https://aip.scitation.org/doi/10.1063/1.434593
- Kang Byung-Nam. "Complex Network Science (Asan Foundation Research Report No. 124)", Seoul : Asan Foundation. 2011. www.asanfoundation.or.kr
- Kim Mun-Jo. "IT and New Order Formation (21st Century Korea Mega Trend Series II)", Gwacheon : Information and Communication Policy Institute, 2005.
- Kim Yong Hak, Park Se-Woong, Jeon So-Young. "An Analysis of the Structure of Online Social Network : Focusing on the Network of", Information Policy, Vol. 13, No. 4, pp.167-185. 2006. https://www.nia.or.kr/site/nia_kor/ex/bbs/List.do?cbIdx=65684
- Park Kwang-Soon. "Analysis on the Hyperlink of News Articles on the Internet Media : Focusing upon the Naver, Daum, Yahoo Site", Journal of the Korea Contents Association, Vol. 10, No. 11, pp.329-340. 2010. http://koreacontents.or.kr/ https://doi.org/10.5392/JKCA.2010.10.11.329
- Park Han-woo. "International Scientific and Scholarly Communication Networks on World Wide Web", Journal Of The Korean Society For Library and Information science, Vol.37, No.2, pp.153-168. 2003. https://kslis.jams.or.kr/co/main/jmMain.kci https://doi.org/10.4275/KSLIS.2003.37.2.153
- Park Han-woo. Nam In-Yong. "Examining the role of perceived website credibility in hyperlinking. Information Policy", Vol.11, No.2, pp.18-34. 2004.
- Song Hae-Ryong. "Media business market and ecosystem", Seoul : Sungkyunkwan University Press. 2010.
- Ahn Min-Ho. "Complexity science and television audience behavior", Media & Society, 17(3), pp.116-150. 2009. https://future.internetsociety.org/introduction-drivers-of-change-areas-of-impact/areas-of-impact/media-society/
- Ye Sung-Ho. "Study on Hanzhong between the dissemination of information on the structure of the 'small world (Small world)'", Chinese and Chinese Studies, 20, pp.33-65. 2013.
- Yoon Young-Soo, Chae Seung-Byung. "Introduction to Complex Systems". Seoul : Samsung Economic Research Institute. 2005.
- Lee Myung-Jin, Jang Ahn-Sic, Park Ki-Tae, Han Sang-Won. "Changes in complex systems and network society (Studies in South Korea Broadcasting Communication Paradigm Shift 08-22)". Gwacheon : Information and Communication Policy Institute. 2008.
- Chang Duk-Jin, Lim Dong-kyun. "Complexity and social structure", In Min Byung-Won, Kim Chang-wook. (Eds.), Complex Workshop. pp.35-91. Seoul : Samsung Economic Research Institute. 2006.
- Jung Ha-Woong, Kim Dong-Sub, Lee Hae-Woong. "Google God knows everything", Seoul : Science Books. 2013.
- Cho, Hang-Hyun, Chae Seung-Byung. "Complexity and Its Applications", In Min Byung-Won, Kim Chang-wook. (Eds.), Complex Workshop, pp.587-622. Seoul : Samsung Economic Research Institute. 2006.
- Ackland, R. "WWW Hyperlink Networks". In D. Hansen, B. Shneiderman & M. A. Smith (Eds.), Analyzing Social Networks with NodeXL: Insights from a Connected World, pp.181-199. Burlington, MA: Morgan Kaufmann. 2010. http://voson.anu.edu.au
- Alexander, J. E., & Tate, M. A. "Web wisdom: How to evaluate and create web page quality", Mahwah, NJ: Lawrence Erlbaum Associates. 1999. https://dl.acm.org/citation.cfm?id=553758
- Arthur, W. B. "Complexity and the economy". Science, Vol. 284, No. 5411, pp.107-109. 1999. http://science.sciencemag.org/content/284/5411/107 https://doi.org/10.1126/science.284.5411.107
- Barnes, S. B. "Computer-mediated communication: Human-to-human communication across the Internet (1st ed.)", 2003. Kwon Sang-Hee. (Translation). Cyber communication theories. Seoul : Sungkyunkwan University publication. 2007. https://book.naver.com/bookdb/book_detail.nhn?bid=3081276
- Bonchev, D., & Buck, G. A. "Quantitative measures of network complexity". In D. Bonchev & D. H. Rouvray (Eds.), Complexity in chemistry, biology, and ecology. pp. 191-235. New York, NY: Springer Science+Business Media. 2005. https://link.springer.com/chapter/10.1007%2F0-387-25871-X_5
- Bonchev, D., & Trinajstic, N. "Information theory, distance matrix, and molecular branching", The Journal of Chemical Physics, Vol. 67, No. 10, pp.4517-4533. 1977. https://aip.scitation.org/doi/abs/10.1063/1.434593
- Chang, T. K., Himelboim, I., & Dong, D. "Open global networks, closed international flows world system and political economy of hyperlinks in cyberspace", International Communication Gazette, Vol. 71, No. 3, pp.137-159. 2009. https://journals.sagepub.com/doi/abs/10.1177/174804850 8100910
- Desikan, P., Srivastava, J., Kumar, V., & Tan, P. N. "Hyperlink analysis-techniques & applications", Stanford, CA: Army High Performance Computing Research Center. 2002. https://core.ac.uk/download/pdf/20699423.pdf
- Gell-Mann, M. "The quark and the jaguar: Adventures in the simple and the complex (reprint ed.)", New York: Owl Books. 1995. https://www.amazon.com/Quark-Jaguar-Adventures-Simple-Complex/dp/0805072535
- Jackson, M. H. "Assessing the structure of communication on the World Wide Web". Journal of Computer-Mediated Communication, Vol. 3, No. 1, 1997. https://academic.oup.com/jcmc/article/3/1/JCMC311/4584371
- Kung, L. "Strategic management in the media: Theory to practice". 2008. Choi Sung-Beom. (Translation). Media Strategic Management, Seoul: Communication Books. 2013. https://lib.dankook.ac.kr/search/detail/CATTOT000001195263
- Milgram, S. "The small world problem". Psychology Today, Vol. 1, No. 1, pp.61-67. 1967. http://mills-soc116.wikidot.com/notes:milgram-small-world
- Nelson, T. H. "Complex information processing: A file structure for the complex, the changing and the indeterminate". In Proceedings of the 1965 20th national conference, pp.84-100. 1965. https://dl.acm.org/citation.cfm?doid=800197.806036
- Park, H. W. "Examining the determinants of who is hyperlinked to whom: A survey of webmasters in Korea". First Monday, Vol. 7, No. 11, 2002. http://firstmonday.org/issues/issue7_11/park/index.html.
- Park, H. W. "Hyperlink network analysis: A new method for the study of social structure on the web". Connections, Vol. 25, No. 1, pp.49-61. 2003. https://www.researchgate.net/publication/200772676_Hyperlink_network_analysis_A_new_method_for_the_study_of_social_structure_on_the_web
- Simon, H. A. "Near decomposability and complexity: How a mind resides in a brain", In H. J. Morowitz & J. L. Singer (Eds.), The mind, the brain, and complex adaptive systems, pp.25-43. Reading, MA: Addison-Wesley. 1995. http://digitalcollections.library.cmu.edu/awweb/awarchive?type=file&item=34196
- Singer, J. L. "Mental processes and brain architecture: Confronting the complex adaptive systems of human thought". In H. J. Morowitz & J. L. Singer (Eds.), The mind, the brain, and complex adaptive systems, pp.1-9. Reading, MA: Addison-Wesley. 1995. https://philpapers.org/rec/MORTMT
- Wardrip-Fruin, N. "What hypertext is". In Proceedings of the 15th ACM conference on Hypertext and hypermedia. pp.126-127, 2004. https://dl.acm.org/citation.cfm?id=1012844
- Wei, C. Y., Evans, M. B., Eliot, M., Barrick, J., Maust, B., & Spyridakis, J. H. "Influencing web-browsing behavior with intriguing and informative hyperlink wording", Journal of Information Science, Vol. 31, No. 5, pp.433-445, 2005. https://www.researchgate.net/publication/220195607_Influencing_web_browsing_behavior_with_intriguing_and_informative_link_wording https://doi.org/10.1177/0165551505055703
- Whitesides, G. M., & Ismagilov, R. F. "Complexity in chemistry". Science, Vol. 284, No. 5411, pp.89-92. 1999. http://science.sciencemag.org/content/284/5411/89 https://doi.org/10.1126/science.284.5411.89