Final Project

edit

Social media analytics is an interdisciplinary area that is used in social science and computer sciences interchangeably. Social media analytics provide a human trace to the social scientist which could be used in wide spectrum of disciplines such as sociology, political sciences, and geology. Social media provides two broad contexts from social scientist perspective; it provides a wide range of data in already well established social science subjects such as political sciences and sociology, and social media sometimes is seen as a fundamental change in underlying assumptions of the social theory. Political scientists can follow unfolding political protest online[1] and the exchange of information between communities of different languages.[2] Meanwhile, it is very difficult to connect the social scientific understanding of social to social media data. For example, the concept of conventional friendship hardly applies to the concept of friendship in social media.[3]

Process of social media analytics

edit

There are three main steps in social analyzing social media: data identification, data analysis, and information interpretation. The preferred way to maximize the value derived at every point during the process, analysts may define a question to be answered. In attempting to analyze the question, analysts may think like detectives, always asking the important questions; “Who? What? Where? When? Why? and How?” These questions help in determining the proper data sources to evaluate, which can greatly affect the type of analysis that can be performed.[4]

Data identification

edit

Data identification is the process of identifying the subsets of available data to focus on for analysis. The data by itself is useless unless it's interpreted, once we start analyzing the data it begins to become useful as it conveys a message. Any data that conveys a meaningful message becomes information. On a high level, unprocessed data takes the following forms to translate into exact message: noisy data; relevant and irrelevant data, filtered data; only relevant data, information; data that conveys a vague message, knowledge; data that conveys a precise message, wisdom; data that conveys exact message and reason behind it. To derive wisdom from an unprocessed data, we need to start processing it, refine the dataset by including data that we want to focus on, and organize data to identify information. In the context of social media analytics, data identification means “what” content are we interested in, in addition to the text of content, we want to know: who wrote the text? Where was it found or on which social media venue did it appear? Are we interested in information from a specific locale? When did someone say something in social media?[4]

Attributes of data that need to be considered are as follows:

  • Structure: Structured data is a data that has been organized into a formatted repository - typically a database - so that its elements can be made addressable for more effective processing and analysis. The unstructured data, unlike structured data, is the least formatted data.[5]
  • Language: Language becomes significant if we want to know the [[sentiment of a post rather than number of mentions.
  • Region: It is important to ensure that the data included in the analysis is only from that region of the world where the analysis is focused on. For example, if the goal is to identify the clean water problems in India, we would want to make sure that the data collected is from India only.
  • Type of Content: The content of data could be, Text; written text that is easy to read and understand if you know the language, Photos; drawings, simple sketches, or photographs, Audio; audio recordings of books, articles, talks, or discussions, Videos; recording, live streams.
  • Venue: The social media content is getting generated in a variety of venues such as news sites, social networking sites (e.g. Facebook, Twitter). Depending on the type of project the data is collected for, the venue becomes very significant.
  • Time: It is important to collect data is posted in the time frame that is being analyzed.
  • Ownership of Data: Is the data private or publicly available? Is there any copyright in the data? These are the important questions to be addressed before collecting data.
 
[6] Social media analytics process

Data analysis is the set of activities that assist in transforming raw data into insight, which in turn leads to a new base of knowledge and business value. In other words, data analysis is the phase that takes filtered data as input and transforms that into information of value to the analysts. Many different types of analysis can be performed with social media data. The data analysis step begins once we know what problem we want to solve and know that we have sufficient data that is enough to generate a meaningful result. How can we know if we have enough evidence to warrant a conclusion? The answer to this question is; we don't know. We can't know this unless we start analyzing the data. While analyzing if we found the data isn't sufficient, reiterate the first phase and modify the question. If the data is believed to be sufficient for analysis, we need to build a data model.[4]

Developing a data model is a process or method that we use to organize data elements and standardize how the individual data elements relate to each other. This step is important because we want to run a computer program over the data; we need a way to tell the computer which words or themes are important and if certain words relate to the topic we are exploring.[7]

In the analysis of our data, it’s handy to have several tools available at our disposal to gain a different perspective on discussions taking place around the topic. The aim here is to configure the tools to perform at peak for a particular task. For example, thinking about a word word cloud, if we take a large amount of data around computer professionals, say the "IT architect", and built a word cloud, no doubt the largest word in the could would be "architect". This analysis is also about tool usage. Some tools may do a good job at determining sentiment, where as others may do a better job at breaking down text into a grammatical form that enables us to better understand the meaning and use of various words or phrases. In performing analytic analysis, it is difficult to enumerate each and every step to take on an analytical journey. It is very much an iterative approach as there is no prescribed way of doing things.[4]

The taxonomy and the insight derived from that analysis are as follows:

  • Depth of Analysis: Simple descriptive statistics based on streaming data, ad hoc analysis on accumulated data or deep analysis performed on accumulated data. This analysis dimension is really driven by the amount of time available to come up with the results of a project. This can be considered as a broad continuum, where the analysis time ranges from few hours at one end to several months at the other end. This analysis can answer following type of questions:
    • How many people mentioned Wikipedia in their tweets?
    • Which politician had the highest number of likes during the debate?
    • Which competitor is gathering the most mentions in the context of social business?
  • Machine Capacity: The amount of CPU needed to process data sets in a reasonable time period. Capacity numbers need to address not only the CPU needs but also the network capacity needed to retrieve data. This analysis could be performed as real-time, near real-time, ad hoc exploration and deep analysis. Real-time analysis in social media is an important tool when trying to understand the public’s perception of a certain topic as it unfolding to allow for reaction or an immediate change in course. In near real-time analysis, we assume that data is ingested into the tool at a rate that is less than real-time. Ad hoc analysis is a process designed to answer a single specific question. The product of ad hoc analysis is typically a report or data summary. A deep analysis implies an analysis that spans a long time and involves a large amount of data, which typically translates into a high CPU requirement.[4]
  • Domain of Analysis: The domain of the analysis is broadly classified into external social media and internal social media. Most of the time when people use the term social media, they mean external social media. This includes content generated from popular social media sites such as Twitter, Facebook and LinkedIn. Internal social media includes enterprise social network, which is a private social network used to assist communication within business.[8]
  • Velocity of Data: The velocity of data in social media can be divided into two categories: data at rest and data in motion. Dimensions of velocity of data in motion can answer questions such as: How the sentiment of the general population is changing about the players during the course of match? Is the crowd conveying positive sentiment about the player who is actually losing the game? In these cases, the analysis is done as arrives. In this analysis, the amount of detail produced is directly correlated to the complexity of the analytical tool or system. A highly complex tool produces more amounts of details. The second type of analysis in the context of velocity is a analysis of data at rest. This analysis is performed once the data is fully collected. Performing this analysis can provide insights such as; which of your company’s products has the most mentions as compared to others? What is the relative sentiment around your products as compared to a competitor’s product?[4]

Information interpretation

edit

Up to this phase, all the steps discussed were to analyze social media content. The important logical step-by-step discussion of what needs to be done in performing analysis; from posing a question, to gathering and cleaning data and then on to doing the analysis. The insights derived from analysis can be as varied as the original question that was posed in step one of analysis. At this stage, as the nontechnical business users are the receivers of the information, the form of presenting the data becomes important. How could the data make sense efficiently so it could be used in good decision making? Visualization (graphics) of the information is the answer to this question.[9]

The best visualizations are ones that expose something new about the underlying patterns and relationships contain the data. Exposure of the patterns and understating them play a key role in decision making process. Mainly there are three criteria to consider in visualizing data.

  • Understand the audience: before building the visualization , set up a goal, which is to convey great quantities of information in a format that is easily assimilated by the consumer of information. It is important to answer, who is the audience, and can you assume the audience has the knowledge of terminologies used. An audience of experts will have different expectations than a general audience; therefore, theexpectations have to be considered.[10]
  • Set up a clear framework: the analyst needs to ensure that the visualization is syntactically and semantically correct. For example, when using an icon, the element should bear resemblance to the thing it represents, with size, color, and position all communicating meaning to the viewer.[10]
  • Tell a story: analytical information is complex and difficult to assimilate, thus, the goal of visualization is to understand and make sense of the information. Storytelling helps the viewer gain insight from the data. Visualization should package information into a structure that is presented as a narrative and easily remembered. This is important in many scenarios when the analyst is not the same person as a decision-maker.[10]

Following are some of the common graphics that are used for visualization of information:

  • Pie charts: pie charts are best used to illustrate the breakdown of a single dimension as it relates to the whole. For example, to depict the number of posts of let’s say 10 users in 24 hours.
  • Bar charts: bar charts are useful for comparing groups of data.
  • Line charts: line charts work best for continuous data, the data those changes over time.
  • Scatter plots: scatter plots can be used to depict a trend or the direction of the data.[4]

References

edit
  1. ^ SEGERBERG, ALEXANDRA; BENNETT, W. LANCE (2011). "Social Media and the Organization of Collective Action: Using Twitter to Explore the Ecologies of Two Climate Change Protests" (PDF). The Communication Review. 14: 197–215.
  2. ^ Bruns, Axel; Burgess, Jean; Highfield, Tim (2013). "The Arab Spring and Social Media Audiences: English and Arabic Twitter Users and Their Networks". www.sagepub.com. Queensland University of Technology, Brisbane, Australia. Retrieved 2016-11-01.
  3. ^ Tinati, Ramine; Phillippe, Olivier; Pope, Catherine; Carr, Leslie; Halford, Susan (2011). "Challenging Social Media Analytics: Web Science Perspectives". ACM: 3–4.
  4. ^ a b c d e f g Ganis, Matthew; Kohirkar, Avinash (2015). Social media Analytics: Techniques and insights for Extracting Business Value Out of Social Media. New York: IBM Press. pp. 40–137. ISBN 978-0-13-389256-7.
  5. ^ "What is structured data? - Definition from WhatIs.com". WhatIs.com. Retrieved 2016-12-06.
  6. ^ Ganis, Matthew; Kohirkar, Avinash (2015). Social media Analytics: Techniques and insights for Extracting Business Value Out of Social Media. New York: IBM Press. pp. 247–248. ISBN 978-0-13-389256-7.
  7. ^ Chekanov, Sergei V. (2016). Numeric Computation and Statistical Data Analysis on the Java Platform. Springer. ISBN 978-3-319-28531-3.
  8. ^ Kitt, Denise (2012-05-24). "Enterprise Social Networks Explained". CRM Switch. Retrieved 2016-11-05.
  9. ^ Steele, Julie (2012-02-15). "Why data visualization matters". O'Reilly Media. Retrieved 2016-12-11.
  10. ^ a b c "The Three Elements of Successful Data Visualizations". Harvard Business Review. Retrieved 2016-12-11.