Social Network Analysis of Fragmented Internet using Twitter Media

Social Network Analysis of Fragmented Internet using Twitter Media

Metric based Social Network Analysis on Internet Fragmentation using Some Popular Twitter Hashtag Networks

Janani Bharath

Dept. of Information Science and Engineering

B.M.S. College of Engineering Bengaluru, India

Mahima A Shejwadkar

Dept. of Information Science and Engineering

B.M.S. College of Engineering Bengaluru, India

R Ashok Kumar

Dept. of Information Science and Engineering

1. College of Engineering Bengaluru, India

   Abstract – Fragmented Internet or Cyber-balkanization is gaining its recognition in recent years because of its cap-potential to take Spiritual pollution in control, due to non accessibility to global data, services and media. Restricting access to the world wide internet has proven advantageous to the government to set limits on virtual activities pertaining to the local laws and regulations, yet makes the mini network absolutely isolated leading to issues with interoperability, business exchanges, collaborative storages causing sink down in certain global markets like cryptocurrency, cloud vendors, gaming businesses and advertising industries. A Twitter analysis using NodeXL on fragmented web, was carried out to understand the information flow on the topic in the form of tweets, posts using hashtag networks and discover peoples influence and perception on internet fragmentation.

   Keywords – Cyberbalkanization, Splinternet, Twitter, Fragmented Internet, NodeXL, Great Firewall, Future of the Internet.

   1. INTRODUCTION

      The word "Internet" was coined in 1974 to describe a global computer network that allows for connectivity and information sharing using standard protocols [1]. Despite the fact that the global ecosystem intends to create an efficient platform for consumers to access data from across the world, some countries refuse to embrace the Data for All concept. External threats, political unambiguity, aggressive national policies, censorships, and trade disputes, to name a few, all contributed to this situation [2]. As a result, in some parts of the World, the Fragmented Internet has emerged. To spell out in another way, the Global internet is divided into mini networks on the basis of geography where the citizens can only access data and services that their national authorities permit [3].

      The 'Great Firewall', implemented in China to regulate internet usage domestically, is one noteworthy example study. This set of legislative regulations was implemented [4] to limit access to specific foreign websites that could create

      societal harm or hatred, as well as to regulate cross-border internet traffic.

      In Russia, internet fragmentation is mainly relied upon in the ongoing Russia-Ukraine war [5]. The Federal government has restricted Twitter and Facebook in Russia since March 2022 in order to prevent the spread of anti-war propaganda and fake news.

      This paper aims at studying the peoples notion on adding borders to the borderless web and the impact of web fragmentation on the current scenario of the Russia-Ukraine war. In this paper, Twitter media is used to analyze the tweets associated with internet splintering using the hashtags

      #FutureOfTheInternet [6], #splinternet [7] and

      #GreatFirewall [8]. The analysis on the number of tweets, users and influencers has been made using NodeXL tool.

   2. REVIEW OF LITERATURE Anderson [9] concluded that though The Great Firewall

      has isolated China from the Global internet, the Chinese government is always in a conundrum when it comes deciding to choose between integrating with growing new technologies and to refrain from these technologies that challenge their traditional censorship regulations.

      The new Sovereign Internet Law implemented in Russia from 2019 to 2021 is nothing different from isolating their network from the Global internet. These amendments were brought into action in order to combat external threats, keep borders under surveillance and create a Centralized Internet System manageable under the State Authority. Epifanova [10] however recommends that Germany and the EU ought to effectively advance the upsides of the world wide web and include significant stakeholders and business methods to maintain furthermore, upgrade its future. Preferably, they ought to foster a typical long haul technique for protecting the web in its current, non-fragmented, genuinely world wide web, which would include broadening the extent of existing stages like the United Nations Internet Governance Forum.

      With respect to interoperability issues of IoT and Fragmented Internet, M.Aly [11] say that the M2M or H2H applications of IoT is severely backed off due to the fragmented internet which leads the IoT industries to face

      difficulty in integrating the diverse service and applications of the IoT objects. Also, the reuse of data becomes questionable with loosely connected networks.

   3. TWITTER ANALYSIS OF THE HASHTAGS AND RESULTS

      NodeXL (Network Overview for Discovery and Exploration in Excel) is a compelling and user-friendly interactive visualization tool used for network analysis [17]. It's a well-structured word excel workbook template with various worksheets that's supposed to represent a network graph. Finding significant "players" in social media networks is one of SNA's key characteristics. Centrality is the term used to pick out significant vertices in a network graph in accordance with ranking, which is generated by values [12]. In-degree centrality is the number of nodes or accounts which points inwards to a node in consideration. In this context, in- degree is used as a popularity metric [13]. The number of directed links designated from the Tweeter is referred to as out-degree centrality. The major influencers in the network are those Tweeters who have the high value of out-degree. Another centrality metric to explore from the standpoint of social network theory is betweenness centrality. The term " Betweenness centrality" refers to how frequently a vertex appears on the shortest path existing between two other vertices [14]. The bridges in the network are the Tweeters with the highest betweenness centrality. The average distance between a vertex and every other vertex in a social network is referred to as closeness centrality [15]. Low closeness centrality is when the Tweeter is directly in link to, or "just a hop away" from, the most of other vertices in the graph network, assuming vertices can only send messages to or affect their current linkages (vertices) [14]. The eigenvector centrality network metric is calculated considering not merely the degree metric, but also the connections of the vertex [16].

      Data Collection

      Data is collected from twitter by considering 3 most trending hashtags on fragmented internet which are

      #FutureOfTheInternet, #splinternet and #GreatFirewall.

      Data Analysis

      The collected twitter hashtags are analyzed using NodeXL operations such as preparation of data, merging duplicate edges, auto filling of columns, sorting the vertices, creating subgraphs, using dynamic filters and changing the layout options are performed to get a network graph.

      Case 1: Using #FutureOfTheInternet

      An analysis of the hashtag which is

      #FutureOfTheInternet is discussed here. The total number of vertices for this network are 1032 and the total number of edges are 1273. The relationship between the nodes is through the mentions or replies to or tweets. Figure 1 depicts a networkgraph for the same.

      Figure 1: A network graph for #FutureOfTheInternet

      In the figure 1, the node with high betweenness centrality is taken into consideration. In this network, user @iingwen has the highest betweenness centrality 3,95,772.935. Hence, if @iingwen is removed from the network then the users connected to him are completely isolated from the rest of the network. The network analysis of this node is described in table 1.

      Table 1: Node Analysis of @ iingwen

      Graph Metrics	Value
      Degree Centrality	277
      In-degree	276
      Out-degree	1
      Betweenness Centrality	3,95,772.935
      Closeness Centrality	0
      Eigenvector Centrality	0.032
      Clustering Coefficient	0

      The user @iingwen joined twitter on 15th June 2010. The total number of followers for this user is 19,08,369 whereas he follows 25 people. The total number of tweets made by the user on twitter is 4,003 tweets.

      Case 2: Using #splinternet

      The total number of vertices for this network are 44 and the total number of edges are 47. The relationship between the nodes is through the mentions or replies to or tweets. Figure 2 depicts a network graph for the same.

      Figure 2: A network graph for #splinternet

      In the figure 2, analyzing the node with betweenness centrality, user @akjaiii has the highest betweenness centrality 30. Hence, if @akjaiii was removed from the network then the users connected to him are completely isolated from the rest of the network. The network analysis of this node has been described in table 2.

      Table 1.2: Node Analysis of @akjaiii

      Figure 3: A network graph for #GreatFirewall

      In figure 3, considering the node with high betweenness centrality, user @langstrumpfpipo has the highest betweenness centrality 156. Hence, if @langstrumpfpipo was removed from the network then the users connected to him are completely isolated from the rest of the network. The network analysis of this node has been tablulated in table 3.

      Graph Metrics	Value
      Degree Centrality	13
      In-degree	13
      Out-degree	0
      Betweenness Centrality	156
      Closeness Centrality	0.077
      Eigenvector Centrality	0.071
      Clustering Coefficient	0

      Table 1.3: Node Analysis of @langstrumpfpipo

      Graph Metrics	Value
      Degree Centrality	6
      In-degree	6
      Out-degree	0
      Betweenness Centrality	30
      Closeness Centrality	0.167
      Eigenvector Centrality	0
      Clustering Coefficient	0

      The user @akjaiii joined twitter on 30th August 2017. The total number of followers for this user is 11,315 whereas he follows 2043 people. The total number of tweets made by this user on twitter is 42,720 tweets.

      Case 3: Using #GreatFirewall

      The total number of vertices for this network are 32 and the total number of edges were 33. The relationship between the nodes was through the mentions or replies to or tweets. Figure 3 depicts a network graph for the same.

      The user @langstrumpfpipo joined twitter on 26th August 2019. The total number of followers for this user is 1,726 whereas he follows 1,524 people. The total number of tweets made by this user on twitter is 27,039 tweets.

   4. CONCLUSION AND FUTURE WORK Betweenness centrality, as previously said, relates to how

frequently a vertex or in this case a user appears on the shortest distance between two other vertices/users in a network. The highly correlated users for each hashtag are discovered using the twitter network analysis with betweenness centrality as the key metric and their influence on other nodes was being inferred using the eigenvector centrality value. Despite the fact that the fragmented internet is thriving in the current environment of network privatization attempts, the analysis shows that social media has yet to take a position encouraging or opposing internet fragmentation. The main cause is considered to be a lack of understanding of the internal framework, as well as the benefits and drawbacks of a fragmented web.

Further, additional research can be done to determine whether fragmentation is beneficial by analyzing real-time survey data.

REFERENCES

[1] Internet,https://en.wikipedia.org/wiki/Internet#:~:text=In%201974% 2C%20Vint%20Cerf%20and,25%20public%20data%20networks,

Accessed on 7th May 2022 at 20:00 IST

[2] SPLINTERNET,https://www.iasgyan.in/daily-current- affairs/splinternet#:~:text=A%20fusion%20of%20the%20words,nati onal%20security%2C%20religion%20and%20more, Accessed on 7th May 2022 at 20:20 IST

[3] Splinternet: The fragmented future of internet? https://www.telecomreview.com/index.php/articles/reports-and- coverage/5060-splinternet-the-fragmented-future-of-internet,

Accessed on 7th May 2022 at 20:35 IST

[4] Free speech vs Maintaining social cohesion, https://cs.stanford.edu/people/eroberts/cs181/projects/2010- 11/FreeExpressionVsSocialCohesion/china_policy.html, Accessed on 7th May 2022 at 20:52 IST

[5] Fortune,https://fortune.com/2022/03/22/russia-war-ukraine-great- firewall-splinternet-internet/, Accessed on 7th May 2022 at 21:00 IST

[6] #FutureOfTheInternet on Twitter, https://twitter.com/search?q=%23futureoftheinternet&src=typeahead

_click&f=top, Accessed on 8th May 2022 at 19:40 IST

[7] #splinternet,https://twitter.com/search?q=%23splinternet&src=recen t_search_click, Accessed on 8th May 2022 at 19:55 IST

[8] #GreatFirewall,https://twitter.com/search?q=%23greatfirewall&src= typed_query&f=top, Accessed on 8th May 2022 at 20:20 IST

[9] Daniel Anderson, Splinternet Behind the Great Firewall of China: Once China opened its door to the world, it could not close it again,

ACM Queue, Vol. 10, Issue 11, November 2012

[10] Epifanova, A. Deciphering Russias Sovereign Internet Law: Tightening Control and Accelerating the Splinternet, Deutsche Gesellschaft für Auswärtige Politik e.V. (DGAP Analysis, 2), 2020

[11] M. Aly, F. Khomh, Y. Guéhéneuc, H. Washizaki and S. Yacout, "Is Fragmentation a Threat to the Success of the Internet of Things?",

IEEE Internet of Things Journal, vol. 6, no. 1, pp. 472-487, Feb. 2019 [12] Wang, Z., Tan, Y., Zhang, M.: Graph-based recommendation on social networks. In: 2010 12th International Asia-Pacific Web

Conference, pp. 116122. IEEE, April 2010

[13] Miller, P.R., Bobkowski, P.S., Maliniak, D., Rapoport, R.B.: Talking politics on Facebook: network centrality and political discussion practices in social media. Polit. Res. Q. 68(2), 377391 ,2015

[14] Hansen, D.L., et al.: Do you know the way to SNA? A process model for analyzing and visualizing social media network data. In: 2012 International Conference on Social Informatics, pp. 304313. IEEE

,2012

[15] Struweg, I.: # Liberty breach: an exploratory usage case of NodeXL Pro as a social media analytics tool for Twitter. In: ICMBD Conference Proceedings, pp. 153163, 2018

[16] Lohmann, G., et al.: Eigenvector centrality mapping for analyzing connectivity patterns in fMRI data of the human brain. PLoS One 5(4), e10232, 2010

[17] NodeXL, https://www.smrfoundation.orgnodexl/, Accessed on 9th May 2022 at 20.00 IST