Dissertation Chapter: Twitter Network Analysis

Date:  2021-06-16 14:45:35
8 pages  (2087 words)
Back to categories
logo_disclaimer
This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.
University/College: 
Wesleyan University
Type of paper: 
Dissertation chapter
logo_disclaimer
This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.

B. Introductions

As we already know that twitter is one of the most microblogging service which still have a lot of users up until today. The service has been massively evolved and expanded if we compare it to the first 2 years of its operational time, it happened because twitter have some new features that interest people to make account and those features are playing important role to make all of user become more and more convenient when they open the service. Without any doubt, we can say that billions of information has been transmitted and accessed through the service and by knowing this, a lot of researcher wants to explore more about the network that can be found in it. But before we talk about some networking aspects of twitter, its better if we get to know about some facts of the service itself.

Twitter start the service around 2006 and up until the beginning of 2017, number of user has been reached more than 300 million and there are at least 500 million of tweets that has been produced each day, which are massive and in fact all of it happened just under one service of social media network. Another fact about twitter is there are 100 million of user which are active daily and 37% of the user have age between 18 years old, meanwhile 29,25% are in the range of 30-49 years old. In terms of demographic, there are 67 million and 13 million of user are in United States and United Kingdom respectively. But, there are 79% of user from overall amount who are not located in US which has been ranked into top 3 countries which have largest amount of user outside of US are Brazil, Japan and Mexico which have 27,7 million, 25,9 million and 23,5 million respectively. Fact that also quite interesting is one country which have highest percentage of internet user and active on twitter is Saudi Arabia [1].

After knowing some aspects of information regarding twitter, surely, we also will find out that this service has sophisticated type of network whether its about between one user to another user, each tweet with the user, one tweet to another tweet or even network between the same hashtag which been used in the different tweets. Mainly there are 3 different connections that clear enough to be explore. Those three network consists of retweet network, mention network and hashtag network. Each of this network have their own characteristic and there are several methodologies that can be used to explore them. In this research, we will try to use and explore more about retweet network. The reason why we are going to talk about retweet network is because it has more clear way to determine each of topic which are being discussed in the timeline and also the retweet network can make sure existing network which will tell us about the connection or network between one user and another. Usually what happen in the real situation is, if there are one user who retweet another users tweet, then automatically there will be connection between those two users and furthermore, we can also know about how close the connection between them by looking up to the frequency of retweet that has been done. Another thing that potentially extracted by the retweet network is we can easily know what are the main interest of each of user just by accessing the contents or what are the information that contained in the tweet which has been retweeted by each twitters user. In order to get more things regarding of information that can be extracted from the network, there are some of research that already done before and one of them is Topic-Sensitive PageRank or also called as TSPR [2]. This term is one of initial way that able to combine between each topic that being discusses in twitter with the structure of network that it have.

At the beginning of 2017 in Saudi Arabia there is one new initiation called as kyf nkon kdo that have to be applied in the societies and among all of the citizens. This term has been initiated by Prince Khalid Al-Faisal which is one member of royal family in Kingdom of Saudi Arabia. Basically, the main purpose of making this initiative is to encourage people to become role model in the real-life situations. So everyone who can become the role model just in one specific action or even in several kind of situation and place, will be called as person who try to achieve the objective that comes from the term. Moreover, this term is also one good way to increase the amount of kindness for all people who lives in Saudi Arabia. As we already know based on the previous paragraph, citizens of Saudi Arabia are really active on social media, especially Twitter. By looking at this fact, in the further sections of this paper we will try to identify and also analyze the retweet network based on the term kyf nkon kdo, then there also some study about communities which founded in the network as well.

In this research, we will try to cover majority part of social network analysis such as degree distributions, social network model, betweenness and centrality of nodes, and also all things that have connection to the community that exist in the network. To make our analysis more reliable and easy to do, we decide to use one of major application which can deal with network analysis called as Gephi. This application is usually used by network analyst to visualize and apply some model to the network that has been generated from some sources and because we will talk about social network analysis about twitter, then all required data or network that will be used in for our analysis has been generated from it. The reason why we use Gephi, because it has many kind of features that can help us to get all of our objectives without sacrificing a lot of time just to get the best modelling for the network, or even to get all precise results of calculations which are very important to this research. More than that, the application also have one good feature called as data laboratory and it have one good functionality that can show the data of network that has been generated or crawled from twitter and it also have user friendly interface. So, all of details regarding to the data can be captured by accessing the data laboratory.

As mentioned above, we are going to solve some problems regarding to social network analysis of retweet network and we use the term kyf nkon kdo as our foundation to get required data to be analyze. Particularly we will try to answer some problems or questions regarding about degree distribution, centrality, betweenness, amount of communities in the network and also, its modularity. Then at the end we will try to prove our hypothesis about important nodes in the network and what are the reason that made those nodes important. Not only that, we will also going to find out some information about each important node in the twitter and try to deliver it by giving some brief explanation to readers with one purpose in mind that is by finding all important aspects of social network that based on retweet network, we just not contributing in field of researcher, but by giving the result through this paper, at the same time we can open another opportunity to other fellow researcher since there are more other terms which still consider new and still need to be explore. In the best of our knowledge, up until this study are being held, there are no such a research that has been done to cover retweet network that based on term kyf nkon kdo and because this is only happen in Saudi Arabia, so not many researcher that located outside of the country are interested in it. Why we choose this term, because for us who live inside the kingdom and as researcher, we think this is one of our activity which can be given as contribution in developing educational area of the country itself. More than that, we also think that the terms is quite interesting to be explored because just by crawling data of tweets or retweets that have one specific topic such as the term, we can extract a lot of information and of course from those information we can get detailed structures in terms of network of the twitter service. Moreover, by doing analysis for it we can assure that there are some nodes that have important role in the network and then also find out and identify whether there are several aspects which can affects those nodes so they can become important in the network or not. Based on these reasons, we think that this research has to be done and its really important to us, since the term kyf nkon kdo can give big impacts in social life and also will definitely change how the people acts in doing their daily activities. As mentioned before, because the term is quietly new and just got initiated at the beginning of 2017, so there is no specific study about it and actually there ra onge specific thing that can possibly make this research a little bit hard to do is the majority of tweets about the topic and has been crawled are written in Arabic language. Surely it will affect our understanding about the contents but in fact, this is one of good challenge to be work on.

In the previous paragraph, we mentioned about proving our hypothesis regarding the retweet network that we are going to discussed in this paper. There is one assumption that pop-up in our mind at the planning phase of this research, we pay more attention to the probability of node (user account) that become important in the network and what are cause(s) that might affect in the process of information distributions along the network or communities within. So, after thinking about that specific condition, we are going to prove that our assumption about the more edges or connections which connected to one node, then probability of that node to become an important node in the network or community become higher or in more simple words, each node which have large amount of in-degree or connection to it, will become important node in the network.

This paper is structured as follows. Section C, is an overview about our background to do the research and there also some related work that have to be reviewed before we perform our research. Section D, contains explanations about data that we use in this research including the source of it and surely all attributes that exist in the data itself. At the same section, our brief explanations about experiments that we have done to the data and also all features in Gephi that we have used to get our objective in this research. Section E, consists of descriptions for all result that we get after finishing our experiment and then some discussion about the result are included in the section too. At the end part, we wrap up this paper with our conclusion and future work in Section F.

C. Background and Related Work

There are several things that made us decide to this research, the first thing is because we want to have more deep knowledge about social network field and get to know each process that happen inside one microblogging service such as twitter. At the same time, we also want to implement our basic knowledge of network analysis such as find and define the degree distribution for one network, then finding communities in it and we also want to get the hands-on experience of implementing some theoretical parts that we have already learn before regarding social network analysis. In this section, there will be some of explanation about theoretical aspects and materials that we already learn and try to do implementation for each of them through data of retweet network that we have. Another important part of research that we will try to deliver in this section is related work. Because as we know, there are a lot of research that has been done and have specific topic in it, so in order to make our research different with the previous work we have to find related work which can be made as our parameter when doing our research and the more important thing is by doing the related work, we will make sure that...

logo_essaylogo_essay

Request Removal

If you are the original author of this essay and no longer wish to have it published on the ProEssays website, please click below to request its removal: