代写大数据分析作业 Social Network Analysis

Social Network Analysis

(with some clarifications)

代写大数据分析作业 which says there is an edge between a node with ID1 and node with ID2. Finally, there is a single “ego” node that is connected to

If you managed to get developer access to Twitter API (procedure below) and would like to build your own network use the following instructions. Otherwise please follow Option 2 or Option 3 instead.

In order to be able to download Twitter data you have to sign up first for Twitter (if you do not have an account already) and then apply for a developer API access at

https://developer.twitter.com/en/apply-for-access

Select Research / Education as the purpose and say in the description: “I am a student at the University of Liverpool taking Big Data Analytics module. I would like to get developer access to Twitter API to do a partial analysis of a small fragment of Twitter network to establish what kind of network it is (scale-free,   random etc.).”  代写大数据分析作业

and in the analysis part: “I would like to get developer access to Twitter API to do a partial analysis of a small fragment of Twitter network to establish what kind of network it is (scale-free, random etc.).” Just select Yes in “Analyze Twitter data” and No elsewhere. More detailed instruction with (a bit outdated) screenshots on CANVAS (Learning Resources).

In order to retrieve data from Twitter you should first read the following guides:

https://towardsdatascience.com/how-to-download-and-visualize-your-twitter-network-f009dbbf107b

and/or

pages 64-68 and 79-82 of Data Science and Complex Networks book (check Learning Resources on CANVAS). Example Python code will also appear there.

Option 2. Otherwise, you can simply download one of the social networks available at:  代写大数据分析作业

https://snap.stanford.edu/data/#socnets

For instance you can use:

https://snap.stanford.edu/data/ego-Facebook.html

Note that this tgz archive contains 10 different “ego” networks: different files with .edges extension. Pick any of these networks (but preferably not too big and not too small).

Each line of such a .edges file looks like:

ID1 ID2  代写大数据分析作业

which says there is an edge between a node with ID1 and node with ID2. Finally, there is a single “ego” node that is connected to all the remaining nodes; please add that node to your network as well.

1. (Option 1)Extract any part of the Twitter social network (an edge from X to Y exists if X follows Y; make sure this network is connected)

(Option 2) Load any social network that you downloaded

(Option 3) Randomly generate a social network using preferential attachment algorithm given in Network Science lecture.  代写大数据分析作业

Ideally your network should have between 100 and 1000 nodes. Describe which option you picked and present how you did that including the most interesting lines of your code with explanation what they do. [10 points]. You can get full marks for each option. (Originally, more points were supposed to be given for option 1, but this was changed due to Twitter rejecting many of your developer account applications. Of course, if Twitter granted you dev access, ideally you should pick the more challenging option 1 to learn more.).

2. Please visualise this network using your favourite graph drawing Python interface (e.g, using   https://networkx.org/documentation/stable/reference/drawing.html, or https://igraph.org/python/, or (typically best looking graphs) https://www.malinga.me/networkx-visualization-with-graphviz-example/)and include this figure in your submission. Try to do your best when plotting this network but of course for big networks one expects a bit of clutter. However, we expect that at least some individual nodes and edges will be visible in such a plot. [20 points] 代写大数据分析作业


3. Plot the chart showing the in-degree of each user in this network: on the X-axis should be in-degree (using logarithmic scale) and y-axis should be p_k= the probability that a user has in-degree k (using logarithmic scale). [20 points]


4. Please analyse the network and identify what kind of network it is: random with no preferential attachment, random with sublinear preferential attachment, scale-free, or hubs-and-spokes topology or fully connected, etc. with a suitable justification [50 points]. This justification should be around half an A4 page long. (hint: look at page 54 of the Network Science lecture slides. You can but do not have to calculate alpha and gamma parameters here.)

更多代写：cs代写   经济代考   会计代写      计算机科学代写  维克技术社区学院代写