As discussed in detail in Chapter 1, Getting Started with R and Social Media Analytics (see A typical social media analytics workflow), we defined some key steps involved in the analysis of data from different social networks. Continuing with the same theme for Twitter, the different use cases we will work on in the next sections can also be broken down into the following key steps:

The data access step involves understanding the APIs and their corresponding R packages to tap into the social network. We've already talked about creating a Twitter app and did a quick connect and extraction of tweets using R. Each of the following sections will make use of the same initial step for data access and then build upon them based on the requirements. We will discuss and provide details for each of the other steps in this workflow as we progress with the use cases. Stay tuned.

Twitter is a speedy medium. Information (along with rumors and nonsense) travels at breakneck speeds across the social network/world. It has now become a norm for an event or news to break first on Twitter and then on any other source of information. It is commonly observed that TV news channels usually play catchup with Twitter during any news breaks. Such a quick spread of information has its own pros and cons, but a discussion of these are out of the scope of this book.

It would be safe to say that if there's anything trending in the connected world, it will be on Twitter first. From brand promotions, sports events, government decisions, election results to news about terror attacks, natural disasters and the notorious fake celebrity death news, Twitter has it all.

Twitter uses search terms, what are called hashtags in Twitter-verse. Any word which begins with a # is termed as a hashtag and instantly becomes searchable on the platform. This not only helps users search...

Twitter timelines are the new battlegrounds for brands, fans and organizations to fight it out and present a winner. Twitter is also a place where users usually rant about their disappointments or share their happiness. The dynamics of human interaction and our urge to share opinionated views on wide ranging topics, from cat pictures to wars and everything in between, have reached an altogether different level.

With its 300 million plus users and counting, Twitter is a virtual country in itself! Its huge user base which generates tweets (or opinions) by the count of millions every minute present a unique opportunity to study and utilize human sentiment and/or opinions. This study of our sentiments and emotion carries a lot more value than just pure academic research (which is, of course, still required by any standards). It carries a lot of business value for companies, governments and celebrities alike.

Before we dive into implementation details and a particular use case...

So far, we have analyzed Twitter and tweets to uncover some interesting insights using techniques and concepts of trend analysis and sentiment analysis. We've utilized different attributes of tweets, like creation time, location and even the text itself, to answer certain questions. In this section, we will touch upon Twitter's network aspects. #BraceYourSelves

A social network is a network or a graph at its core. In formal words, a social network is generally a graph representing its users as nodes (or vertices) linked to each other based on certain relationships called edges. Each social network has its own definition of these relationships. For this section, we will focus on Twitter's relationships and network in general.

In Twitter-verse as we all know, there are no friends! A friend relationship is usually a bidirectional relationship, that is, if A is a friend of B, then it is safe to say that B is also a friend of A (well, usually; see Facebook...

A tweet is far more than just 140 characters, and Twitter offers quite a lot to play with for a social network. We covered a lot of ground in this chapter by looking at many concepts and solving use cases based on real Twitter data. We learned about different Twitter objects and its APIs. We created an app of our own and utilized R's twitteR package to connect and tap into its APIs. We performed trend analysis to understand how a hashtag is used by tweeple and its temporal affects. We also solved a use case involving sentiment analysis. Through this use case, we first understood the key concepts related to sentiment analysis and then employed them to understand what emotions @POTUS conveys through his tweets. We also performed hierarchical clustering of tweets to visualize common themes using a dendrogram. The final use case analyzed Twitter from a network/graph analysis stand point. We utilized R's different libraries to prepare a network map of followers and perform...