Semester 2 BUS5CA Social Media Analysis Report: Case Studies

This report presents a social media analysis conducted for a customer analytics course, focusing on understanding customer preferences and sentiments through two case studies. The analysis involves exploring the impact of online advertising and communication on customer engagement, particularly examining the relationship between keywords, shares, and sentiments across different topic categories. Case Study A utilizes R programming to analyze the relationship between dependent and independent variables, considering factors like article properties and channel types. Task 1 involves splitting the dataset, sorting based on shares, and performing linear regression. Task 2 delves into keyword analysis using text mining techniques in SAS, including preprocessing, text processing, and feature generation. The report concludes with findings on the impact of social marketing, emphasizing the use of R for analyzing client data and identifying influences on share counts. The report references various research papers related to sentiment analysis and statistical analysis methods.

Customer Analytics and Social Media

Table of Contents
1. Introduction.........................................................................................................................................1
2. Case study A........................................................................................................................................1
Task 1 – Exploration of Article Properties..............................................................................................1
Results of Task 1.................................................................................................................................2
Task 2 – Keyword Analysis using text miner..........................................................................................9
Results of Task 2.................................................................................................................................9
3. Conclusion...........................................................................................................................................9
4. References...........................................................................................................................................9
1. Introduction
This report discusses the social marketing analysis conducted in the client organization. This research is
intended to know about the impact of social marketing in the client organization. This report is assessing
the impact of article properties. This process will be done using R programming (Review of Dual
Sentiment Analysis, 2015). Here the client company dataset will be analyzed to find out the influence of
different variables on the number of shares.
2. Case study A
Task 1 – Exploration of Article Properties
Here the task is to identify the relation between the dependent variables and independent variables. In this
task first, the given dataset is split into six datasets based on the data channel type. Then the dataset is
again modified. Initially sorted the dataset based on the number of shares (Higher number to lower
number)
Then the top 10 % values are taken for the analysis. Then the dataset is analyzed using the following R
code. The r code carries out the statistical analysis on the uploaded dataset (Sentiment Analysis - A
Review, 2015) (Yang, 2012). And find the linear regression between the dependent variable and
1

independent variable. And finally, the linear regression results are plotted as a scatter plot. And the code
is shown in the below given figure.
Code for upload the dataset
Code for statistical calculation and plotting results
Results of Task 1
Lifestyle
Import Dataset
Shares vs tokens title
Shares vs token content
Shares vs links
2

Share vs images
Shares vs videos
Shares vs weekend
Entertainment
Import Dataset
Shares vs tokens title
Shares vs token content
3

Shares vs links
Share vs images
Shares vs videos
Shares vs weekend
Bus
Import Dataset
Shares vs tokens title
4

Shares vs token content
Shares vs links
Share vs images
Shares vs videos
Shares vs weekend
Socmed
Import Dataset
Shares vs tokens title
5

Shares vs token content
Shares vs links
Share vs images
Shares vs videos
Shares vs weekend
World
Import Dataset
6

Shares vs tokens title
Shares vs token content
Shares vs links
Share vs images
Shares vs videos
Shares vs weekend
Tech
7

Import Dataset
Shares vs tokens title
Shares vs token content
Shares vs links
Share vs images
Shares vs videos
Shares vs weekend
8

Task 2 – Keyword Analysis using text miner
Steps involved in text mining using SAS is explained in this section. At first, the test dataset and training
datasets are downloaded and preprocessed. Then the text processing process is initiated. At this stage, the
semantic text analysis process is carried out. Then the feature generation step is carried out (Statistical
Analysis of Medical Data Using SAS, 2006) (ZHAO, QIN and LIU, 2010). After that data mining and
patter discover the process is carried out. In this section, the insights of the data set can be identified. But
we need to carry out visualization for better understanding. And this is the last step in this process. Here
the developed results are illustrated.
Results of Task 2
Overview of the most common key words founded from the dataset is illustrated in the above figure.
3. Conclusion
Social marketing analysis in the client organization has completed successfully. From the conducted
analysis we can know about the impact of social marketing in the client organization. And the findings of
the conducted analysis has explained in this report. In the first part of the report assessing the impact of
the article, properties have been explained (The Statistical Analysis of Series of Events, 2014). This
process has been done using R programming. Here the client company dataset has been analyzed to find
out the influence of different variables on the number of shares.
9

4. References
Review of Dual Sentiment Analysis. (2015). International Journal of Science and Research
(IJSR), 4(11), pp.2323-2326.
Sentiment Analysis - A Review. (2015). International Journal of Science and Research (IJSR),
4(12), pp.1842-1845.
Statistical Analysis of Medical Data Using SAS. (2006). Biometrics, 62(2), pp.634-635.
The Statistical Analysis of Series of Events. (2014). Springer Verlag.
Yang, J. (2012). Using SAS for data management, statistical analysis and
graphics. Pharmaceutical Statistics, 11(4), pp.346-346.
ZHAO, Y., QIN, B. and LIU, T. (2010). Sentiment Analysis. Journal of Software, 21(8),
pp.1834-1848.
10