Facebook Network Analysis (for a facebook page) using Gephi - A students' guide

Students who are pursuing their course on Social Network Analysis in some form or the other, can use Gephi, a free software which can be used to visualize and manipulate graphs. Gephi is an interactive visualization and exploration platform for all kinds of social / other networks and complex systems.It can also process dynamic and hierarchical graphs. It runs on various platforms which include Microsoft Windows, Linux and Mac OS X. LinkedIn maps is also based on Gephi tool with some modifications.

Some of the applications of Gephi are:

• Exploratory Data Analysis
• Link Analysis
• Social Network Analysis
• Biological Network analysis
• Poster creation

Some of the technological aspects of Gephi are:

• Ergonomic interface: It is based on NetBeans UI
• High-performance: It has a built-in 3D rendering engine
• Native file formats: GDF (GUESS), GraphML (NodeXL), GML, NET (Pajek), GEXF, etc are some of the file formats which are supported by Gephi
• Customizable by plugins: layouts, metrics, data sources, manipulation tools, rendering presets, etc.

Now, it is very easy to use Gephi to do data analysis of facebook data for bipartite network. The following steps may be used to do the above mentioned analysis:

1. Login to Facebook through an user account & ‘like’ the facebook page for which analysis needs to be done.
2. Using the Netvizz app (http://apps.facebook.com/netvizz/) on Facebook extract a dataset for a bipartite network containing users and posts
3. Download 1 gdf format file and 2 tsv format files (total 3 files)
4. Open the gdf format file using Gephi and analyze
5. Open the tsv format file using Microsoft Excel

Using Gephi's layout tool, statistics tool and filter tool, various analysis on degree, betweenness and closeness centrality, network density, path length, diameter, HITS, modularity, clustering coefficient, etc can be obtained.

Some Modularity analysis on Gephi can be shown as below:

Using Big Data in Telecom - how it can change the game play

Increasing ARPU (Average Revenue Per User), reducing customer churn (subscriber base) and reducing OPEX (Operational Expenses) are three very important priorities for any CSPs (Communication Service Providers) or Telecom Service Providers in general. With the rise in the usage of telecommunication devices in our day-to-day lives, the amount of digital content is increasing tremendously in the data networks. 

Having good insights from the big data, utilizing the insights to ramp up the revenue streams, catering to the customer complaints more effectively, taking feedback for the new service offerings and improving upon them, etc can really create differentiation among the competitors in the telecom industry. Big telecom companies, around the world have realized this opportunity and are taking good amount of initiatives to harness their big data analysis to their help.

As per a study done by IBM, successfully harnessing big data can help the telecom companies to mainly achieve the following three critical objectives:

1. Deliver smarter services that generate new sources of revenue and content across the network
2. Transform internal and external operations to achieve business and service excellence across the different SBUs
3. Build smarter CDN and data networks to drive consistent, high-quality customer experience

There are many service sector companies which are now offering the following service packages to the the telecom operators (The list is not exhaustive):

• Proactive call centers
• Smarter and effective marketing campaigns
• Network Analytics
• Location based services

Some examples of big data analytics in work for the telecom companies and their success stories are:

• A Tier-1 Mobile Service Provider, based out of the US, transforms call centers operation with real time access to Product and services data (customer/users)
• Globe Telecom, based in the Philippines, used big data analytics to improve the effectiveness of their promotions by 600%
• Ufone, based out of Pakistan is deploying big data analytics to improve marketing "offer acceptance rate" by 25%
• XO Communications, based out of US, used big data analytics to improve customer experience and an annual OPEX reduction of USD 15 million per year
• T-Mobile AG, Deutsche Telekom (Germany) uses big data to optimize network performance & reduce OPEX costs
• Sprint, based out of the US, leverages big data analytics to put real-time intelligence into the network driving a 90% increase in network capacity

So it is quite evident that companies across the world are increasing depending on the big data analytics for their improvement and this trend is opening up different avenues for data scientists and big data analysts.

Image source: Internet

Towards Big data in a big way - an introduction

In the modern times, people all over the world are increasingly creating digital information through their usage of cellphones, digital cameras, social media websites like Twitter, LinkedIn, Facebook,etc. The digital data deluge is on the rise at a very fast pace. Online streaming of data is a very common thing to do over the internet, which primarily needs huge data transfers across networks. It has been estimated that almost 90% of the digital data residing in the world of ours was created in the past two years itself. Also most of this data (85%) is of the unstructured kind of data.

Few things which might not be known to few, regarding digital data are:

1) Internet has a huge memory and it literally forgets nothing, absolutely nothing - anything and everything that gets created digitally, gets stored in the internet, either through different storage mediums or in raw form as a unstructured stream of ones and zeroes (which can be retrieved as meaningful information when needed) in the network/ memory of the network elements.

2) Privacy issues are increasing with the addition of different internet applications - Marketers and security agencies are extremely interested to monitor practically everyone for their ulterior motives. NSA has been allegedly reported to be following secretive practices to keep record of all the online activities of world citizens over the internet. E-commerce websites use the information content of the web browser cache to understand the consumer behavior and consumer buying pattern.

3) All the big data are getting analysed and collected by different companies at various geographies and capacities - with the increasing demand for analytics among the corporate and industry trend analysts, different companies are increasingly looking for different ways to understand their customers better. While social media websites are a rich source of such information, other Web 2.0 based websites are also important.

With the growing reach of internet and related technology, more and more data will be generated in future and analyzing such data can help the companies to reap tremendous amounts of profits. Data scientists are in huge demand now-a-days and in a way they hold the key to success for many businesses.

Big data have three unique properties of Volume, Variety and Velocity. Along with this time-stamp and location information adds meaning to the data. Analysis of big data is gaining importance primarily because of the fact that unlike analysis of sample based data, analysis on big data takes into effect the entire set of data (as is) and hence all sampling errors can be avoided.

Photo courtesy: intellihq

लाखों बकवास कर लूँगा मैं...

उनकी एक मुस्कुराहट के लिए , लाखों बकवास करता हूँ मैं 

काश उनसे कह पाता की उनको कितना प्यार करता हूँ मैं 

यूँ तो बातें हो जाती हैं हमारे बीच 

पर शायद मेरे बोल कम पड़ जाते हैं 

चित्त तो अब परेशान ही रहता है 

दिल की बात सुनने से भी अब डरता हूँ मैं 

उनकी एक मुस्कुराहट के लिए साला अब भी मरता हूँ मैं !

सयाने से इस जहान में बस उनका साथ मिले तो

बेबस इन निगाहों को उनके आने का इशारा मिले तोह 

मुरझाये से इस चेहरे को हंसने का एक बहाना मिले तो 

स्थिर पड़े इस मन को उड़ने का एक बहाना मिले तो 

वो हँसे, ऐसा कुछ कहने के  लिए अल्फाज़ अगर मिल जाये 

लाखों बकवास कर लूँगा मैं ...प्यार मेरा मुझे जो मिल जाये !


Transliteration in English:

unki ek muskurahat ke liye, lakhon bakwas karta hoon main
kaash unse keh pata ki unko kitna pyar karta hoon main

yun toh baatein ho jati hain humare beech
par shayad mere bol kam pad jatey hain
mann to ab hamesha pareshan hi rehta hai
mann ki baat ko sunne se bhi ab darta hoon main

unki ek muskurahat ke liye sala ab bhi marta hoon main !

sayane se is jahan mein bas unka saath mile toh
bebas in nigahon ko unke aane ka ishara mile toh
murjhaye se is chehre ko hansne ka ek bahana mile toh
sthir pade is mann ko udne ka bahana mile toh

wo hanse, aisa kuch kehne ke liye alfaz agar mil jaye
lakhon bakwas kar loonga main...pyaar mera mujhe jo mil jaye !