Session 1

Lambton College
School of Computer Studies
AML-3204 Social Media Analytics

The concept of Social Media
Session-1
2
Objectives
After the completion of this lecture you will learn about:
◇Evaluation and Marking Scheme
◇The Internet and WWW
◇Social Media
◇Usage of Social Media in Business and Marketing Purposes
◇Social Media Issues
◇Data Analytics
◇Social Media Analytics
◇Importance of Social Media Analytics
2
Evaluation
1. Evaluation methods
Evaluation Method Weight
Tests (2 @ 20%) 40%
Assignments (2 @ 10%) 20%
Final Term Project (1 @ 40%) 40%
Total 100%
2. Marking Scheme
Mark(%) Grade Mark(%) Grade
94-100 A+ 67-69 C+
87-93 A 63-66 C
80-86 A- 60-62 C-
77-79 B+ 50-59 D
73-76 B 0-49 F
70-72 B-
3
4
The Internet
◇The Internet is a global web of computers connected to each other by
different networking media
◇It is the largest network in the world
◇It is the super information highway
◇Routers are used to connect different networks
◇No one owns the Internet
◇No organization formally manages it
◇TCP/IP protocols are used to control the communication in Internet
4
5
The Usage of Internet

◇Send/Receive emails
◇Send/Receive files between computers
◇Setup online meetings
◇Participate in discussion groups
◇Web surfing
◇Online Gaming
◇Research
◇News Channel
5
6
The World Wide Web
(WWW)
◇WWW consists of information organized into Web pages
containing text and graphic images.
◇Different web resources are identified by Uniform Resource
Locators (URLs).
◇A website is a collection of linked web pages that has a
common theme or focus.
◇The main page is called the home page.
◇Websites are accessed using companies LAN or using an
Internet Service Provider
6
7
Different Terminologies
◇Browser
◇Search Engine
◇URL
◇Domain
◇Hypertext Markup Language
◇Client/Server Architecture
◇IP Addressing
◇HTTP
◇HTTPS
7
8
What is Social Media?

◇Social media use interactive technologies to create or
exchange information
◇Content is created, distributed and shared by individuals on the
web
◇Users can provide feedback on the shared contents
◇It is broad
◇It has fewer limitations on what someone can share
◇Different means of mass communications like newspapers,
radio stations, and television channels have more limitations on
what they can publish
8
9
Benefits of Social Media
◇Companies can reach a wider audience fast

because social media reaches users in a shorter
amount of time
◇Social Media can facilitate collaboration
◇Easy to distribute information
◇Strengthens and builds relationship
◇Engages different parties
9
10
Dos of Social Media

◇Add values
◇Be authentic
◇Know your audience
◇Respect your audience
◇Listen to others
◇Have your strategy ready
◇Be creative
◇Think thoroughly before you post
10
11
Types of Social Media

◇Blogs: Tumblr
◇Microblogs: Twitter
◇Collaborative Projects: Wikipedia
◇Media Sharing Networks: Instagram, Snapchat, YouTube
◇Discussion Forums: Reddit
◇Social Networks: Facebook
◇Consumer Review Networks: Yelp
◇Interest-based networks: Goodreads
11
12
Issues of Social Media

◇Security Issues
Malicious Banner
Adware
Phishing Attack
◇Cyberbullying
◇Image-based abuse
◇Offensive or illegal content
◇Adverse health risks of Internet addiction
◇Sleep problems due to use of electronic media, etc.
12
13
What is Data?
◇Data is the raw material
◇Data is transformed into information and ultimately knowledge
◇Value creation:
13
14
Data Analytics
◇It is a process of inspecting, cleansing, transforming, and modeling data with the goal of
discovering useful information, suggesting conclusions, and supporting decision making
14
15
Social Media Analytics

◇The art and science of extracting valuable hidden insights from
vast amounts of semi-structured and unstructured social media
data to enable informed and insightful decision making
◇We have enormous amount of social data:
500 million tweets per day
95 million photos and videos are uploaded to Instagram per day
500 hours of video are uploaded to YouTube per day
350 Million photos are uploaded per day
15
16
Social Media Analytics Tasks

◇Platform Analytics
Tracks impression
Engagement with the post using likes, retweets, shares,
comments, etc.
◇Monitoring Social Media
Monitor mentions of brand, hashtag, keyword, etc.
◇Measuring Sentiment
Measures sentiment of your brand
Measures the sentiment of a topic which is related to your
company
16
17
Social Media Analytics Tasks..

◇Measuring Conversations
Can provide insights into the success of the campaign
◇Analyzing Images
Image analytics is applied to capture the information given on the images
◇Understanding Audiences
To find out what group to target
To determine contents for the messages for different groups
To determine campaign type that will increase the engagement
You want to find the answers of the questions: who is discussing your brand?,
What is the age/gender?, etc.
17
18
Social Media Analytics Tasks..

◇Influencer Analytics
 Here we want to find our who is advertising your brand in a
positive way with the audiences
◇Social Media Intelligence
We can use monitoring and analytics to discover insights from
the data
We can answer different strategic questions, such as
audiences engagement with a topic
18
19
Conclusion
We have studied
◇The evolution of Internet and WWW
◇Characteristics of Social Media
◇Usage of Social Media for Business
◇Data Analytics
◇Social Media Analytics
◇Importance of Social Media Analytics
19
20
Review Questions
◇What are the don’ts of social media?
◇What is the difference between web 1.0 and web 2.0?
◇Compare Facebook, Twitter and Instagram in terms of total
number of users, total posts created/shared, photos uploaded,
etc.
◇ Which networking device do we use to connect different
networks?
◇Who is the owner of the Internet?
◇Which protocol suites define how communication should be
done?
◇What is WWW? 20
21
Review Questions..
◇How is a web resource identified in WWW?
◇What is HTML?
◇What is the difference between HTTP and HTTPS?
◇What are some dos of social media?
◇Which social media is used for Microblogging?
◇Which social media platform is used for discussion
forums?
◇What are the security issues for social media?
21
22
Review Questions..
◇How do you create value from data?
◇What are the steps of data analytics?
◇What are the different social media analytics tasks?
◇What do you use to analyze a social media platform?
◇How do you monitor social media?
◇Why do you want to measure sentiment?
◇Why do you want to understand your audiences?
22
23
Next session..
In the next session we will study the following topic:

Return on Investment for Social Media
23
Lambton College

Return on Investment for Social Media
Session-2
25
Objectives
◇ Social Media Value Creation Model (VCM)
◇ Social Media Return on Investment (ROI)
◇ Social Media Value Metrics to measure ROI
◇ Social Media Engagement (SME) Metrics
◇ Social Media Influence (SMI) Metrics
◇ Social Media Popularity (SMP) Metrics
Labs
◇ BMI Analysis
◇ Normalization
◇ Parsing
25
26
Social Media Analytics Value Creation
Model
◇ Research by MIT Sloan Management found that 67% of the respondents
of a survey reported that by employing analytics their companies
gained a competitive advantage
◇ Another study found that 56% of the marketers do not know how to
incorporate social media with their business outcomes
◇ Creating value with social media analytics entails harnessing cost-
effective and commercially worthy insights from social media semi-
structured and unstructured data that can ideally lead to competitive
advantage
26
27
Model..
The value creation process is done through a set of activities:
◇ Defining what the value is
◇ Aligning the value creation with business objectives
◇ Capturing the value using analytics
◇ Sustaining the value for a long period
◇ As the support activities to the value creation model analytics
infrastructure is used
27
28
Model..
28
29
Value of Social Media

◇ From a business perspective, value includes all sets of tangible and
intangible assets that determine the overall success of a firm in the long run.
◇ Firms engaging through social media expect to get more sales and cost
savings (financial value), brand awareness, and relationship building (non-
financial value)
◇ Customers engaging with firms through social media want to get
discounts and lowering their transaction cost (financial value), and being
part of the brand community (non-financial value)
29
30
The categories of value

The value derived from social media can be categorized into two:
1. Value to Firms:
Tangible V2F
Intangible V2F
2. Value to Customers
Tangible V2C
Intangible V2C
30
31
Social Media Tangible V2F

Tangible or financial V2F from social media can come in a variety of forms:
◇ More sales
◇ Market growth
◇ Cost reduction
Research suggests that positive comments over social media can lead to
significant value
◇ A study showed that an eBay seller could earn an additional $45.76 with
the help of 675 positive opinions
◇ Another study shows that emotional words in social media text (such as
hope, worry, and fear) have a significant impact on share market indices
31
32
Non-tangible V2F
◇ Brand awareness
◇ Brand name
◇ Brand loyalty
◇ Customer engagement
◇ Mass collaboration
◇ Crowd-sourcing
◇ Idea generation
◇ Connectivity
◇ Customer Satisfaction
◇ Website traffic, etc.
32
33
Non-tangible V2C
◇ Product awareness
◇ Brand association
◇ Brand Connectivity
◇ Brand involvement
◇ Service quality
◇ Information quality
◇ Product quality, etc.
33
34
More on Tangible V2F

◇ Brand Sales
◇ Repeat Sales
◇ Market share
◇ Market growth
◇ Repeat volume
◇ Cost savings
◇ Customer acquisition cost
◇ Customer retention cost
◇ Customer lifetime value
34
35
Tangible V2C
◇ Discounts
◇ Competitive price
◇ Group buying
◇ Social buying
◇ Volume discount
◇ Promotions
◇ Low transaction cost
◇ Savings
◇ Easy buying, etc.
35
36
Social Media Return on Investment

◇ Return on Investment (ROI) are the gains from an investment less the
cost of an investment
◇ Firms make financial and/or non-financial investments in social media
◇ Financial investments include: paid social media ads, search engine
optimization, and paid media
◇ Non-financial investments include: Organic SEO and ads
◇ These investments lead to: brand reach, engagement, likes, comments,
clicks, Google ranking, and user-generated content --- these returns are
collectively known as social media returns
36
37
Social Media Return on Investment
37
38
Social Media Value Metrics

◇ One way to gauge social media value or returns on
social media investment is through value metrics or key
performance indicators (KPIs)
◇ KPIs are quantifiable measures used to gauge a
company’s success at reaching targets
◇ The KPIs help a business stay focused on achieving
their goals and objectives
38
39
Value metric: Social Media Engagement
(SME)
◇ It is the overall responsiveness and interaction of a brand with its
customers through social media
◇ It is the communication connections between brands and its stakeholders
through various social media channels
◇ The essential idea of SME is to encourage customers to interact and
share their experiences
◇ Active social media engagement means how many of the total followers
are actively interacting with the content
◇ Since social media platforms vary significantly regarding scope,
engagement, and use, SME metrics must be platform specific
39
40
Social Media Engagement (SME)…

For example, if a brand’s objectives were to measure social media engagement through
Twitter, KPIs for this would be:
◇ Re-tweet %: Re-tweets here are the tweets of the followers re-transmitted by brands
◇ Replies %: percentage of replies in the total tweets of a brand in a particular time
frame
◇ User Mentions: the average number of people mentioned per tweet by a brand.
In the case of Facebook, SME metrics include:
◇ No. of times content generated by brands
◇ No. of times content generated by fans
◇ No. of times content shared
◇ No. of comments posted by fans
◇ No. of replies posted by brands
40
41
Social Media Influence (SMI)

◇ SMI is the capacity or ability of brands to have a positive effect on social media
users, fans, and followers
◇ The positive effect can result in brand loyalty, buying decisions, and positive
behaviors
◇ For Twitter, we can use the following KPIs to measure SMI:
1. Tweets favorited: the proportion of a brand’s tweets favorited by other Twitter
users
2. Re-tweets: the proportion of a brand’s tweets re-tweeted by Twitter users
3. Brand mentions: it is the average number of times a brand is mentioned per
tweet by Twitter users
41
42
Social Media Popularity (SMP)

◇ A brand popularity is defined as the extent to which a brand is widely purchased by
the general public
◇ In social media, popularity is the overall fame or social media share of a brand
◇ Popularity over social media may lead to financial and social returns for a brand
◇ On Twitter, we can measure a brand’s popularity as follows:
◇ Followers: it is the number of fans/supporters of a brand
◇ Ratio of followers: A ratio of more than one means the brand is more popular
◇ Listed: it is the number of people who added a brand to their public list
42
43
Conclusion
We have studied
◇ Social Media Value Creation Model (VCM)
◇ Social Media Return on Investment (ROI)
◇ Social Media Value Metrics to measure ROI
◇ Social Media Engagement (SME) Metrics
◇ Social Media Influence (SMI) Metrics
◇ Social Media Popularity (SMP) Metrics
43
44
Review Questions
◇What activities are done to create value?
◇State an example of a non-financial value for a firm?
◇State an example of a non-financial value for a customer?
◇State an example of a financial value for a customer?
◇What are the categories of value?
◇ State some examples of tangible V2F
◇ State some examples of non-tangible V2F
◇ State some examples of non-tangible V2C
44
45
Review Questions..
◇ State some examples of tangible V2C
◇State the difference between organic and non-organic ads.
◇How do get social medial returns?
◇What could be some SME metrics for Facebook?
◇What could be some SME metrics for Twitter?
◇What is the difference between SME and SMI?
◇Which metrics do we use for SMI measurement?
◇Why do we use SMP for? For Twitter, what metrics can be used to
measure SMP?
45
46
Next session..

Social Media Analytics and Business
Intelligence
46
Lambton College

Social Media Analytics and Business Intelligence
Session-3
48
Objectives
◇ Social Media Analytics
◇ The difference of Social Media Analytics and Business Analytics
◇ Different types of Social Media Analytics
◇ Challenges to Social Media Analytics
Labs
◇ Parsing
◇ Normalization
◇ One Hot Encoding Vector
◇ Veracity Calculation
48
49
What is Social Media Analytics?

Social Media Analytics “is the art and science of
extracting valuable hidden insights from vast amounts of
semi-structured and unstructured social media data to
enable informed and insightful decision making.”
49
50
What is Social Media
Analytics?..
art: It is an art of interpreting and aligning the insights
gained with business goals and objectives.
50
51
Analytics?..
science: It is a science, as it involves systematically
identifying, extracting, and analyzing social media data
using sophisticated tools and techniques.
51
52
Analytics?..
extracting: The data needs to be mined from public
databases
52
53
Analytics?..
extracting valuable hidden insights from vast amounts
of semi-structured and unstructured social media data to
valuable hidden insights: It carries business value and
are not usually visible to the naked eyes.
53
54
Analytics?..
extracting valuable hidden insights from vast amounts
of semi-structured and unstructured social media data
to enable informed and insightful decision making.”
vast amounts of semi-structured and unstructured:
The data has four Vs properties: volume, variety, velocity
and veracity.
54
Copyright © 2018 Gohar F. Khan
55
55
56
What is Social Media Analytics?

enable informed and insightful decision making”
informed and insightful decision making: The objective
of SMA is to make good business decision based on data
56
57
Interest in Social Media Data
57
58
Social Media Analytics vs Business
Analytics
Table 1. Social media vs. conventional business analytics (Khan, 2015)
58
59
SMA related terms (Google
trends)
59
60
Purpose of Social Media Analytics

◇ Measure brand loyalty
◇ Tracking product/service/campaign impact
◇ Business forecasting
◇ Business intelligence and market research
◇ Connecting and engaging with current customers
◇ Finding and engaging with new customers
◇ Getting and analyzing feedback on product/service
◇ Generating business leads
60
61
Types of Social Media Analytics

◇ Descriptive
◇ Diagnostics
◇ Predictive
◇ Prescriptive
61
62
Descriptive Analytics
◇ Descriptive analytics are carried out to answer
questions about events that have already occurred
Sample questions can include:
◇ What was the sales volume over the past 12 months?
◇ What is the number of support calls received as
categorized by severity and geographic location?
◇ What is the monthly commission earned by each sales
agent?
62
63
Descriptive Analytics..
◇ It is estimated that 80% of generated analytics results
are descriptive in nature
◇ Value-wise, descriptive analytics provide the least
worth and require a relatively basic skillset
63
64
Diagnostic Analytics
◇ Diagnostic analytics aim to determine the cause of a
phenomenon that occurred in the past using questions that focus
on the reason behind the event (Root Cause analysis)
Example questions include:
◇Why were a specific sales less than the other sales?
◇ Why have there been more support calls originating from the
Eastern region than from the Western region?
◇Why was there an increase in patient re-admission rates over
the past
64
65
Predictive Analytics
◇ Predictive analytics are carried out in an attempt to determine the outcome of
an event that might occur in the future
◇Predictive analytics try to predict the outcomes of events, and predictions are
made based on patterns, trends and exceptions found in historical and current
data.
Example questions include:
◇ What are the chances that a customer will default on a loan if they have
missed a monthly payment?
◇ What will be the patient survival rate if Drug B is administered instead of
Drug A?
◇ If a customer has purchased Products A and B, what are the chances that
they will also purchase Product C?
65
66
Prescriptive Analytics
◇ It answers the questions of what should we do? Why
should we do it?
◇ It is used when we have important, time-sensitive or
complex decisions to make
◇ For example, to offer incentive to customers who are
likely to leave your business
66
67
Summary: Types of Social Media
Analytics
67
68
Case Study: Item Recommendation

◇ Situation: A customer buys a product at amazon.ca
◇ Task: Recommend other products that this customer
is likely to buy
◇ Amazon uses a technique that is called association
analysis
68
Case Study – Price
69
Prediction
◇ Situation: A person brings her car to a car dealership
to sell
◇ Task: The dealership wants to estimate the price of
the car
◇ We Makecan predict
Model
the Millage
Year
price of theHorsepower
Color
car using the price of
Sold Price
previousToyotacars
Camry 2011 150K White 180 $15,000
BMW X3 2015 25K Blue 270 $35,000
Honda Civic 2005 350K Black 145 $3,000
Ford F150 2007 200K Red 220 $16,000
69
70
Working with Categorical Data: One-hot
Encoding
◇ Many algorithms cannot work with categorical data directly
◇ The categories must be converted into numbers
◇ A one hot encoding is a representation of categorical variables
as binary vectors
◇ This first requires that the categorical values be mapped to
integer values
◇ Then, each integer value is represented as a binary vector
that is all zero values except the index of the integer, which is
marked with a 1
70
71
Example: One-hot Encoding

◇ Assume we have a sequence of labels with the values ‘red’
and ‘green’
◇ ‘red’, ‘red’, ‘green’
◇ 0,0,1
◇ One-hot encoding vector for:
◇ ‘red’: [1,0] and ‘green’: [0,1]
◇ So, ‘red’, ‘red’, ‘green’ becomes:
◇ [1,0][1,0][0,1]
71
72
Conclusion
We have studied
◇ Social Media Analytics
◇ The difference of Social Media Analytics and Business
Analytics
◇ Different types of Social Media Analytics
◇ Challenges to Social Media Analytics
◇ One-hot encoding vector
72
73
Review Questions
◇What are the challenges of Social Media Analytics?
◇What is meant by veracity?
◇When do we use one-hot encoding vectors?
◇What is the difference between SMA and BA in terms of data
type?
◇ What do you want to calculate to find the quality of the dataset?
◇ What type of social media analytics you are going to use to
answer questions such as "What happened?"
◇What type of social media analytics you are going to use to
answer questions such as "Why did it happen?"
73
74
Review Questions..
answer questions such as "What will happen?"
answer questions such as "What do I do?“
◇ What is association analysis?
74
75
Next session..

Social Media Analytics Infrastructure
75
Lambton College

Social Media Analytics Infrastructure
Session-4
77
Objectives
◇ The Seven Layered approach to know your customer
◇ Eight layers of social media analytics
◇ Social Media Monitoring
◇ Social Media Listening
◇ Social Media Analytics Cycle
Labs
◇ Matrix Multiplication
◇ Traditional Matrix Factorization
◇ Neural Embedding based Matrix Factorization
77
78
A Seven Layer approach to Know your
customer
◇ Where are they? Location Analytics
◇ What they say? Text Analytics
◇ What they do? Actions Analytics
◇ What they search? Search Engine Analytics
◇ How they network? Network Analytics
◇ How they navigate? hyperlink Analytics
◇ How they use apps? Apps Analytics
78
79
Eight Layers of Social Media Analytics

Networks
Multimedia Text
Apps 8 Layers of Actions

Social Media
Analytics
Search Hyperlink
Engines s
Location
79
80
Layer One: Text

Social media text analytics deals with the
extraction and analysis of business insights from
textual elements of social media content, such as:
◇ comments
◇ tweets
◇ blog posts and
◇ Facebook status updates
80
81
Layer One: Text

What kind of business insights can you extract
from textual data?
◇ Sentiment mining
◇ Intention mining
◇ Topic and idea mining, etc.
81
82
Layer Two: Network

Social media network analytics extract, analyze,
and interpret personal and professional social
networks, for example, Facebook and Twitter.
What kind of business insights can you extract
from network data?
◇ Influential People
◇ Network Structure
82
83
Twitter Network Example
83
84
Layer Three: Actions

Social media actions analytics deals with extracting, analyzing,
and interpreting the actions performed by social media users,
including likes, dislikes, shares, mentions, and endorsement.
Purpose:
◇ Engagement tracking
◇ Social media marketing campaign’s performances
◇ Understanding brand popularity (likes, views, clicks)
◇ Referral Analytics
84
85
Layer Four: Apps

Apps analytics deals with measuring and
optimizing user engagement with mobile
applications (or apps for short).
85
86
Layer Five: Hyperlinks

Hyperlink analytics is about extracting, analyzing,
and interpreting social media hyperlinks
86
87
Layer Six: Location

Location analytics, also known as spatial
analysis or geospatial analytics, is concerned with
mining and mapping the locations of social media
users, contents, and data.
87
88
Layer Seven: Search Engines

Search engines analytics focuses on analyzing
historical search data for gaining a valuable insight
into a range of areas, including trends analysis,
keyword monitoring, search result and
advertisement history, and advertisement
spending statistics.
88
89
Layer Eight: Multimedia

Multimedia analytics focuses on analyzing the
audio, video and image files shared by the users in
the different social networking platforms.
89
90
Social Media Monitoring

Social Media Monitoring allows you to keep track of:
◇ Every conversation
◇ User accounts
◇ Keyword and hashtag related to your business brand
To define social monitoring we can say:
It is the act of monitoring social media for information relevant to
your business
90
91
Social Media Monitoring tools

◇ Google Alerts
◇ Hootsuite
◇ Talkwalker
◇ Reddit Search
◇ Reputology
◇ Netbase, etc.
91
92
Social Media Listening

Social listening uses monitoring tools to track different social
media platforms for mentions and conversations related to your
brand. Then it analyzes them for insights to discover
opportunities to act.
It’s a two step process:
◇ Step-1: Use monitoring tools to monitor different parameters
related your business
◇ Step-2: Analyze the information for ways to put what you learn
into action. That can be something as small as responding to a
happy customer, or something as big as shifting your entire
brand positioning.
92
93
Social Media Analytics Cycle

Identification
Interpretation Extraction
Business
Objectives
Visualization Cleaning
Analyzing
93
94
Social Media Analytics Process
94
95
Identification
Framing the right question and knowing what data to analyze is
extremely crucial in gaining useful business insights
◇ Searching and identifying the right source of information for
analytical purposes
◇ Type of data needed (text, actions, search engine)
◇ Source of data (Twitter, Facebook)
◇ Public data vs., company owned source
95
96
Extraction
Methods, tools, and skills needed to extract the data.
◇ API based Extraction: In simple words, are sets of
routines/protocols that social media service companies (e.g.,
Twitter and Facebook) have developed that allow users to
access small portions of data hosted in their databases.
96
97
Cleaning
This step involves removing the unwanted data from the
automatically extracted data.
◇ Missing values
◇ Incorrect/false data
◇ Coding (e.g., Gender)
◇ Anonymize
97
98
Analyzing
◇ Depending on the layer of social media analytics under
consideration and the tools and algorithm employed, the steps
and approach to take will greatly vary.
◇ The overall objective at this stage is to extract meaningful insights
without the data losing its integrity
98
99
Visualization
Effective visualization is particularly helpful with complex and
large data sets because it can reveal hidden patterns,
relationships, and trends.
◇ Network data (with whom)
◇Topical data (what)
◇Temporal data (when)
◇Geospatial data (where)
99
100
Conclusion
We have studied
◇ The Seven Layered approach to know your customer
◇ Eight layers of social media analytics
◇ Social Media Monitoring
◇ Social Media Listening
◇ Social Media Analytics Cycle
100
101
Review Questions
◇ What is the seven layer approach to know your customers?
◇What are the eight layers of social media analytics?
◇What kind of business insights can you extract from textual data?
◇What kind of business insights can you extract from network
data?
◇Which layer of social media analytics does engagement tracking?
101
102
Review Questions..
◇ Which layer of social media analytics is known as spatial
analysis?
◇ Which layer of social media analytics includes advertisement
history to complete the analysis?
◇ What is the difference between social media monitoring and
listening?
◇ In social media analytics life cycle, what is done in the
identification step?
102
103
Next session..

Strategy
103
Lambton College

Value Creation Strategy, Data Cleaning and
Visualization
Session-5
105
Objectives
◇ Value creation strategies to get value out of data
◇ Data cleaning techniques
◇Techniques to visualize data
Labs
◇ Traditional Matrix Factorization
◇ Neural Embedding based Matrix Factorization
105
Value creation stages
The value creation steps can be divided into the following nine
stages:
◇Business Case Evaluation
◇Data Identification
◇Data Acquisition & Filtering
◇Data Extraction
◇Data Validation & Cleansing
◇Data Aggregation & Representation
◇Data Analysis
◇Data Visualization
◇Utilization of Analysis Results
106
Value creation stages..
107
108
Data Identification
◇ The Data Identification stage is dedicated to identifying the datasets
required for the analysis project and their sources.
◇ Identifying a wider variety of data sources may increase the probability of
finding hidden patterns and correlations. For example, to provide insight, it
can be beneficial to identify as many types of related data sources as
possible, especially when it is unclear exactly what to look for.
◇ Depending on the business scope of the analysis project and nature of
the business problems being addressed, the required datasets and their
sources can be internal and/or external to the enterprise.
108
109
Data Identification..
◇In the case of internal datasets, a list of available datasets from
internal sources, such as data marts and operational systems,
are typically compiled.
◇ In the case of external datasets, a list of possible third-party
data providers, such as data markets and publicly available
datasets, are compiled. Some forms of external data may be
embedded within blogs or other types of content-based web
sites, in which case they may need to be harvested via
automated tools.
109
110
Data Acquisition and
Filtering
◇During the Data Acquisition and Filtering stage, the data is
gathered from all of the data sources that were identified during
the previous stage.
◇The acquired data is then subjected to automated filtering to
fetch the related data for the analysis
◇ Depending on the type of data source, data may come as a
collection of files, such as data purchased from a third-party data
provider, or may require API integration, such as with Twitter.
110
111
Filtering..
◇ In many cases, especially where external, unstructured data is
concerned, some or most of the acquired data may be irrelevant
(noise) and can be discarded as part of the filtering process.
◇ Data classified as “corrupt” can include records with missing
or nonsensical values or invalid data types. Data that is filtered
out for one analysis may possibly be valuable for a different type
of analysis.
◇ Therefore, it is advisable to store a verbatim copy of the
original dataset before proceeding with the filtering. To minimize
the required storage space, the verbatim copy can be
compressed.
111
112
Filtering..
◇ Both internal and external data needs to be persisted once it gets
generated or enters the enterprise boundary.
◇ For batch analytics, this data is persisted to disk prior to analysis. In
the case of real time analytics, the data is analyzed first and then
persisted to disk.
◇ Metadata can be added via automation to data from both internal
and external data sources to improve the classification and querying
◇ Examples of appended metadata include dataset size and
structure, source information, date and time of creation or collection
and language-specific information.
112
113
Extraction
113
114
Data Extraction
◇ Some of the data identified as input for the analysis may arrive in
a format incompatible with the solution model. The need to address
disparate types of data is more likely with data from external
sources.
◇ The Data Extraction stage, is dedicated to extracting data and
transforming it into a format that the underlying solution can use for
the purpose of the data analysis
◇ The extent of extraction and transformation required depends on
the types of analytics and capabilities of the solution. For example,
extracting the required fields from delimited textual data, such as
with webserver log files, may not be necessary if the underlying
Data solution can already directly process those files.
114
115 Data Validation and
Cleansing
◇ Invalid data can falsify analysis results. Unlike
traditional enterprise data, where the data structure is
pre-defined and data is pre-validated, data for social data
analytics can be unstructured without any indication of
validity. Its complexity can further make it difficult to
arrive at a set of suitable validation constraints.
◇ The Data Validation and Cleansing stage is dedicated
to establishing often complex validation rules and
removing any known invalid data.
115
116 Data Aggregation and
Representation
◇ Data may be spread across multiple datasets,
requiring that datasets be joined together via common
fields, for example date or ID. In other cases, the same
data fields may appear in multiple datasets, such as date
of birth. Either way, a method of data reconciliation is
required or the dataset representing the correct value
needs to be determined.
◇ The Data Aggregation and Representation stage,
shown in the next figure, is dedicated to integrating
multiple datasets together to arrive at a unified view.
116
117 Data Aggregation and
Representation..
117
118
Data Aggregation and
Representation..
Performing this stage can become complicated because
of differences in:
◇ Data Structure
◇ Semantics – A value that is labeled differently in two
different datasets may mean the same thing, for example
“surname” and “last name.”
118
119
Data Analysis
◇ The Data Analysis stage is dedicated to carrying out the actual analysis
task, which typically involves one or more types of analytics.
◇ This stage can be iterative in nature, especially if the data analysis is
exploratory, in which case analysis is repeated until the appropriate pattern
is retrieved.
◇ Depending on the type of analytic result required, this stage can be as
simple as querying a dataset to compute an aggregation for comparison. On
the other hand, it can be as challenging as combining data mining and
complex statistical analysis techniques to discover patterns or to generate a
statistical or mathematical model to depict relationships between variables.
◇ Data analysis can be classified as confirmatory analysis or exploratory
analysis, the latter of which is linked to data mining
119
120
Data Analysis…
120
121
Data Analysis…
◇ Confirmatory data analysis is a deductive approach
where the cause of the phenomenon being investigated
is proposed beforehand. The proposed cause or
assumption is called a hypothesis.
◇ The data is then analyzed to prove or disprove the
hypothesis and provide definitive answers to specific
questions. Data sampling techniques are typically used.
121
122
Data Analysis…
◇ Exploratory data analysis is an inductive approach that
is closely associated with data mining. No hypothesis or
predetermined assumptions are generated.
◇ Instead, the data is explored through analysis to
develop an understanding of the cause of the
phenomenon. Although it may not provide definitive
answers, this method provides a general direction that
can facilitate the discovery of patterns or anomalies.
122
123
Data Visualization
◇ The ability to analyze massive amounts of data and find useful
insights carries little value if the only ones that can interpret the
results are the analysts.
◇ The Data Visualization stage is dedicated to using data
visualization techniques and tools to graphically communicate the
analysis results for effective interpretation by business users.
◇ Business users need to be able to understand the results in
order to obtain value from the analysis and subsequently have
the ability to provide feedback, as indicated by the dashed line
leading from stage 8 back to stage 7
123
124
Data Visualization
◇ The results of completing the Data Visualization stage provide
users with the ability to perform visual analysis, allowing for the
discovery of answers to questions that users have not yet even
formulated
◇ The same results may be presented in a number of different
ways, which can influence the interpretation of the results.
Consequently, it is important to use the most suitable
visualization technique by keeping the business domain in
context.
◇ Another aspect to keep in mind is that providing a method of
drilling down to comparatively simple statistics is crucial, in order
for users to understand how the rolled up or aggregated results 124
125
Utilization of Analysis Results

◇ Subsequent to analysis, results being made available to
business users to support business decision-making, such as via
dashboards, there may be further opportunities to utilize the
analysis results.
◇ The Utilization of “Analysis Results stage” is dedicated to
determining how and where processed analysis data can be
further used to gain advantage
◇ Depending on the nature of the analysis problems being
addressed, it is possible for the analysis results to produce
“models” that encapsulate new insights and understandings
about the nature of the patterns and relationships that exist within
the data that was analyzed. 125
126
Utilization of Analysis Results..

◇ A model may look like a mathematical equation or a set of
rules. Models can be used to improve business process logic and
application system logic, and they can form the basis of a new
system or software program.
◇ Common areas that are explored during this stage include the
followings:
 Input for Enterprise Systems – The data analysis results may
be automatically or manually fed directly into enterprise
systems to enhance and optimize the behaviors and
performance.
126
127
Utilization of Analysis Results..

◇ For example, an online store can be fed processed customer-
related analysis results that may impact how it generates product
recommendations. New models may be used to improve the
programming logic within existing enterprise systems or may
form the basis of new systems.
◇ Business Process Optimization – The identified patterns,
correlations and anomalies discovered during the data analysis
are used to refine business processes. An example is
consolidating transportation routes as part of a supply chain
process. Models may also lead to opportunities to improve
business process logic.
127
128
Data cleaning: How to Handle Missing
Values?
◇ Fill in the missing value manually: tedious + infeasible?
◇ Ignore the tuple containing missing values:
 usually done when class label is missing (assuming the task
is classification)
128
129 Data cleaning: How to Handle Missing
Values?..
◇ Global Estimation
 the attribute mean/median for numeric attributes
 the most probable value for symbolic (i.e. categorical) attribute
◇ Local Estimation:
 the attribute mean/median for all the tuples belonging to the same
class (for numeric attributes)
 the most probable value within the same class (for symbolic
attributes)
129
130
Data cleaning: How to Handle Noisy Data?

◇ There are different techniques to handle noisy data, for
example:
◇ Clustering: Detect and remove outliers (An outlier is a value
that does not follow the general pattern of the rest)
◇ Regression: Smooth by fitting the data into regression
functions
130
131
Cluster Analysis
131
132
Regression Analysis
◇ Fit the data to a function
◇ Data points too far away from the function are outliers.
132
133
Conclusion
We have studied
◇ Value creation strategies to get value out of data
◇ Data cleaning techniques
◇ Techniques to visualize data
133
134
Review Questions
◇How does the process of value creation starts?
◇What is done in the data extraction process?
◇From what stage, we go back to data analysis stage
again?
◇What is a confirmatory analysis?
◇What is exploratory analysis?
◇What is a model?
134
135
Review Questions
◇While classifying, if a class label is missing then what
do we do with that tuple?
◇What is global estimation?
◇What is local estimation?
◇Example of supervised learning technique.
◇Example of unsupervised learning technique.
135
Lambton College

Text Analytics
Session-6
137
Objectives
After the completion of this lecture, you will learn about:
◇ Text Analytics
◇ Types of Social Media Text
◇ Reasons for Text Analytics
◇ Steps of Text Analytics
Classwork:
◇ Please read the paper on Neural Collaborative Filtering
◇ Create a summary slide and present it to the class
Lab: LDA Modeling
137
80% of Data is Unstructured
◇Database notes:
Call center transcripts
Other CRM
◇Email
◇Open-ended survey responses
◇Web pages
◇News Groups
◇Documents themselves
◇Competitive information
◇Reviews, tweets, comments
◇Photos, Videos, info graphics
138
139
What is Text Analytics?

◇ Social media text analytics, also known as text mining,
is a technique to extract, analyze, and interpret hidden
business insights from (unstructured) textual elements of
social media content.
139
140
Type of Social Media Text

◇ Dynamic Text: Dynamic text is a real-time social
media user-generated text or statement to expresses an
opinion about content or information posted over social
media.
◇ Dynamic text is usually smaller in length (e.g., a
couple of sentences), diverse in nature, and is updated or
deleted more frequently
◇ Examples: Tweet, Comments, Discussion,
Conversation, Reviews
140
141
Type of Social Media Text..

◇ Static Text: Static social media text is usually large in
length (e.g., several paragraphs) and is generated,
updated, or deleted less frequently.
◇Examples:
• wiki content
• a blog page
• Word documents
• Corporate reports
• electronic mail (e-mail), and
• news transcripts.
141
142
Text Industry
142
143
Deployment Models
◇ On-premise model
• It is comparatively expensive option, but provides extra security and
control
◇Cloud based model

• Attractive to small and medium scale businesses providing them a
cost effectiveness, scalability, and lower risks solution.
143
144
Key Players
◇IBM (Watson Analytics)
◇Lexalytics
◇Microsoft (Text analytics APIs)
◇SAS Text
◇SAP
144
145
Applications for Text
Analytics
◇ Social media data
◇Surveys
◇‘Reading’ email
◇Call centre data
◇Abstracts
◇Document management
◇Corporate history
◇Scientific publications
◇Thematic understanding of website
◇Database notes
145
146
Purpose of Text Analytics

Sentiment
Analysis
Perpose of
Concept Trends
Mining Social Media Mining
Text Analytics
Intention
Mining
146
147

◇ Sentiment Analysis: Sentiment analysis analyzes and
categorizes social media text (mostly dynamic text) as
being positive, negative, or neutral
147
148

◇Intention Mining:
 Intention or intent mining aims to discover users’
intention (such as buy, sell, recommend, quit,
desire, or wish) from media text.
 Can be used to find new potential customers who
intend to buy a product (or services), and
 service existing customers who have trouble with a
product.
148
149

◇ Trends Mining
 Trends mining, also known as predictive
analytics, is used to predict future events.
 Google Trends
 Vast amount of social media data (e.g.,
comments and tweets) can be mined to identify
patterns and trends for new product
development
149
150

◇ Concept Mining
 Concept mining aims to extract ideas and concepts from
documents.
 Concept mining is useful in extracting ideas from large
amounts of static social media text, such as wiki content,
a web page, Word documents, and news transcripts.
 Concept mining can be employed to classify, cluster, and
rank ideas.
150
Text analytics mechanism

Concept
Maps
Text Attract
Attitudes
Clustering
Grow
Surveys Categoriza-
tion
NLP
Text
Retain
Actions Concepts
Web Trending
Channel
Outcomes
Information
Business UI
Extraction
Operational Fraud
Systems
Attributes
Prediction
Customer Data Expert UI Business
Data Collection User
151 Copyright 2003-4, SPSS Inc.

151
152
Natural Language
Processing
◇ NLP is a field of computer science, artificial
intelligence and computational linguistics concerned
extract meaning from text in a way that is very similar to
how the human brain understands language.
152
153
Supervised vs. Unsupervised Learning

◇ In supervised the algorithm is trained on the type of data it can
expect to analyze.
◇ Unsupervised machine learning method try to discover

patterns rather than trying to fit the data into the predefined
structure.
153
154
Steps in Text Analytics

•Clustering •Dynamic text: tweets,
•Classification comments, reviews
•Association analysis •Static Text: Wiki content,
•Predictive analysis blogs, websites, reports.
•Sentiment analysis
Source
Text Mining Identification
Text Parsing
Text
&
Transformation
Filtering
•Terms count •Stemming
•Frequency count •Parts of speech
•Co-occurence metrics •Named entities
extraction
•Stop words
•Filtering
154
155
Steps in Text Analytics..

◇ Clustering http://tweettopicexplorer.neoformix.com/#n=NYTimes
◇ Clustering or cluster analysis groups objects based on
similarity in non-overlapping groups.
◇
155
156

◇Association
◇ Association or association rule mining is a data-mining
technique used to find frequent patterns, correlations,
associations, or causal structures from data sets.
◇ It can be used, for example, to find that a user who liked a
social media content A and B is 90 percent likely to also like
content C.
156
157

◇ Classification
◇ Is used to find similarities in the document and group them
with predefined labels based on the themes contained in the
document.
◇ For example, an e-mail can be classified as spam based on
its contents.
157
158
Conclusion
We have studied
◇ Text Analytics
◇ Types of Social Media Text
◇ Reasons for Text Analytics
◇ Steps of Text Analytics
158
Lambton College

Recommender Systems
Session-7
160
Objectives
After the completion of this lecture, you will learn about:
◇ What is a recommender system?
◇ How does a recommender system work?
◇ Different types of Recommender Systems
◇ Collaborative Filtering Based Recommender System
◇ Content-based Recommender System
◇ Hybrid Recommender System
◇ Text Mining and Recommender Systems
160
Recommender System
◇ Recommender systems (RS) help to match users with items to
ease information overload
◇ They are information filtering systems that recommend items to
different users based on users’ preferences, interests and behaviors
◇ For example: Related movie recommendation in Netflix
161
Recommender Systems..
◇ Recommender systems reduce information overload by
estimating relevance
162
Types of Recommender Systems
◇ Collaborative filtering (CF) based Recommender System:
Recommends items by identifying other users with similar
interest and it uses a rating matrix
◇ Content based (CT) Recommender System: Recommends
items based on the content similarity. Different features of items
are used as the side information to design the recommender
system
◇ Knowledge-based Recommender Systems: Recommends
items based on explicit knowledge about the items and user
preferences
◇ Group Recommender Systems: Recommends items
depending on the group attachment for a user
◇ Hybrid Recommender System: combines CF and CT based
techniques
163
Collaborative Filtering Based Recommender
Systems
◇ Tell me what's popular among my peers
164
Content-Based Recommender Systems
◇ Show me more of the same what I've liked
165
Knowledge-based Recommender Systems
◇ Tell me what fits based on my needs
166
Group Recommender Systems
◇ Here recommendation is provided depending on the group
attachment for a user
167
Hybrid Recommender Systems
◇ Hybrid recommender systems combine two or more
recommendation techniques in order to increase the overall
performance
◇ The main idea is using multiple recommendation techniques to
suppress the drawbacks of an individual technique in a combined
model
168
Hybrid Recommender Systems..
169
170
User-based collaborative filtering

The basic technique
◇ Given an "active user" (Alice) and an item ‘i’ not yet rated by Alice
◇ The goal is to estimate Alice's rating for this item, by
finding a set of users (peers) who liked the same items as Alice in the past and who have rated item ‘i’
using, e.g., the average of their ratings, to predict if Alice will like item ‘i’
do this for all the items Alice has not seen and recommend the best-rated item
Item1 Item2 Item3 Item4 Item5

Alice 5 3 4 4 ?
User1 3 1 2 3 3
User2 4 3 4 3 5
User3 3 3 1 5 4
User4 1 5 5 2 1
170
171
Types of Feedback
◇ Explicit Feedback:
￮ Explicit ratings (usually 1 to 5) given by users
◇ Implicit Feedback:
￮ Clicks, page views, time spent on page, downloads …
￮ Can be used when explicit feedback is missing or in addition to explicit
ones
￮ Social Network shares (impressions/likes/replies/etc.)
￮ Interaction Data (Comment/Like/Share)
171
172
Calculation of Implicit Feedback
Score
◇ FSui = W L * totalLike + W C * totalComment + W s * totalShare+ W i*
impressions + W tc* totalClickthrough + W rts* reTweetreShareScore + …
◇ Here:
FSui = Feedback score for one user on one item
W i = Weight for impression
W L = Weight for likes
W c = Weight for comments
W s = Weight for shares
W tc = Weight for ClickThrough
W rts = Weight for tweeter retweet and facebook reshare
172
173
Matrix Factorization
◇ Given a n * m matrix R with some entries unknown
o n rows represent n users
o m columns represent m items
o Entry 𝑅𝑖𝑗 represents the i-th user’s rating on the j-th item
◇ We are interested in the unknown entries’ possible values
o i.e., predict users’ ratings
◇ We can model the problem as R = U * V
◇ Here, U and V are two feature matrices for users and items
173
174
Content-based recommendation
◇ Collaborative filtering does NOT require any information about the items.
It just focuses on the Users’ feedback
￮ In our calculation if we can include item’s information, then we would
be able to recommend items based on their contents. If a user likes
Soccer, then we will be able to recommend soccer related posts.
◇ The objective here is:
o To build user’s profile based on the contents of the items the user has
interacted before
o To recommend items those are similar to a user’s profile
174
175
Group Recommender System

◇ Recommenders are usually designed to provide recommendations
adapted to the preferences of a single user u.
◇ However, in many situations the recommended posts are consumed by a
group of users
o A travel with the friends
o A movie to watch with the family during the Christmas holidays
o Music to be played in a car for the passengers
◇ When a user is new, both content-based and collaborative filtering
recommendation don’t work, in this case, we can apply group
recommendation algorithm, recommending an item based on the group
membership of the new user.
175
176
Text Mining
◇ It seeks for novel and useful patterns in the data
◇ It mostly deals with unstructured data: word documents, PDF files, text
excerpts, etc.
◇ To perform text mining - first, impose structure to the data, then mine the
structured data
◇ Benefits of text mining are obvious especially in text-rich data
environments
￮ law (court orders), academic research (research articles), finance
(quarterly reports), medicine (discharge summaries), biology (molecular
interactions), technology (patent files), marketing (customer comments),
etc.
◇ Recommend documents, Spam filtering, e-mail prioritization,
etc. 176
177
Text Representation
◇ A document is represented using a feature vector
◇ A collection of documents is called a corpus
◇ A document is composed of individual tokens or terms
◇ Each document is one instance
◇ Bag of Words:
￮ Treat every document as just a collection of individual words
￮ Ignore word order, grammar, structure of the sentence and punctuations
￮ Each word in the document is treated as a keyword of the document
￮ Each feature is represented by a one or a zero
177
178
Bag of Words example
178
179
Normalized Term Frequency

◇ “computer” appears twice in Doc-201 and this document contains 200 words in total
◇ “computer” appears ten times in Doc-202 and this document contains 2,000,000 words in
total
◇ What is the normalized term frequency for “computer” for Doc-201?
◇ What is the normalized term frequency for “computer” for Doc-202?
◇ Inverse Document Frequency (IDF)
◇ IDF (t) =
◇ The IDF of a term shows how significant that term is in the entire collection of
documents, rarer words get higher IDF score
179
180
TF-IDF
◇ A model to represent a document
◇ TF-IDF (t, d) reflects how important a word t is to a document d in a corpus

◇ Document D501 contains the word “chess” 12 times, and it contains 100 words in total.
There are 10,000,000 documents in total and 300,000 documents contain the word “chess”
◇ So, what is the TF-IDF (“chess”, D501)?
180
181
N-gram Sequences
◇ To preserve word order, we use N-gram sequences
◇ Includes sequences of adjacent words as term. Adjacent pairs are commonly called bi-
grams or 2-grams
◇ Example: “Text mining finds hidden patterns” will be transformed into:
◇ {text, mining, finds, hidden, patterns, text_mining, mining_finds, finds_hidden,
hidden_patterns}
◇ Can you find any 3-grams here?
◇ Do you see any problems here?
181
182
N-gram..
◇ Which one is a bi-gram for the following sentence:
“She walks very fast”
1. {she, fast}
2. {walks, fast}
3. {she, very}
4. {walks, very}
182
183
Profile Similarity For
Recommendation
◇ Now, if a user reads a document d, then we can find the similar documents by using
different similarity finding algorithms. For example, we can use cosine similarity to find the
similar documents and recommends them to the user
◇ Cosine Similarity is a similarity formula to measure similarity of two
posts/articles. It is also known as vector-based similarity.
183
184
Conclusion
We have studied
◇ What is a recommender system?
◇ How does a recommender system work?
◇ Different types of Recommender Systems
◇ Collaborative Filtering Based Recommender System
◇ Content-based Recommender System
◇ Hybrid Recommender System
◇ Text Mining and Recommender Systems
184
185
Review Questions
◇What is the input to a collaborative filtering-based recommender system?
◇What is the input to a content-based recommender system?
◇What is the input to a hybrid recommender system?
◇When is a group recommender system beneficial?
◇What is done to apply matrix factorization to a User-Item rating Matrix R?
◇Difference between explicit feedback and implicit feedback?
◇ Say, you collect comments given by the users on a movie. You calculate
and use the sentiment score as the rating for that movie. Here, this rating is
implicit or explicit feedback?
185
186
Review Questions..
◇What is the difference between N-gram sequences and "Bag of Words"
approach?
◇What is a problem for N-gram sequences?
◇What is a benefit of using N-gram sequences?
◇What does cosine similarity algorithm do?
186
Lambton College

Capturing Value with Network Analytics
Session-8
188
Social Media Network Analytics

◇ Social media network analytics deals with
constructing, analyzing, and understanding social
media networks
◇ NETWORKS ARE the building blocks of social
media and can carry useful business insights
188
189
Purpose of Network Analytics

◇ Understand overall network structure
• number of nodes
• number of links
• density
• clustering coefficient
• diameter
189
190
Purpose of Network Analytics..

◇ Find influential nodes and their rankings
• Degree
• Betweenness
• closeness centralities
◇Find important links and their rankings weight
• betweenness and centrality
◇ Find cohesive subgroups
• pinpointing communities within a network
◇Investigate multiplexity
• analyzing comparisons between different link types, such as friends vs enemies
190
191
Common Network Terms

◇ Network
◇ Social Networks
◇ Social Network Site
◇ Social Networking
◇ Social Network Analysis
191
192
Common Network Terms:
Network
◇ At a very basic level, a network is a group of nodes that
are connected with links
◇ Nodes can represent:
• Individuals
• Organizations
• Countries
• Computers
• Websites, etc.
192
193
Common Network Terms:
Network..
◇ Links represent the relationship among the nodes
• Friendship
• Trade
• Authorship
• Hyperlinks, etc.
193
194
Common Network Terms: Social
Networks
◇ A social network is a group of nodes and links formed by
social entities where nodes can represent social entities
such as people and organizations
◇ Real World Social Networks: A network among classmates
is an example of real-world social network.
◇ Online Social Networks: A Twitter follow-following network is
an example of an online social media network.
194
195
Common Network Terms: Social Network
Site
◇ A social network site is a special-purpose software (or social
media tool) designed to facilitate the creation and maintenance of
social relations
◇ Examples:
• Facebook,
• Google+,
• LinkedIn, etc.
195
196
Common Network Terms: Social
Networking
◇ The act of forming, expanding, and maintaining social
relations is called social networking
◇ Using social network sites, users can, for example, form,
expand, and maintain online social ties with family, friends,
colleagues, and sometimes strangers
196
197
Common Network Terms: Social Network
Analysis
◇ Social network analysis is the science of studying and
understanding social networks
◇ Has root in variety of fields:
• Graph Theory
• Sociology
• Information Science
• Communication Science
197
198
Network Structure
◇ Variety of network structures exist:
• Random network
• Scale-free network
• Centralized networks, etc.
◇We have different structures for network for node degree
distribution
◇ Degree of a node measures the number of links a node has to
the other nodes in a network
198
199
Network Structures based on Degree
distribution
◇ Degree distribution is the probability distribution of
nodes degrees over the whole network
◇ Degree distribution tries to capture the difference in the
degree of connectivity between nodes in a graph
199
200
Random Network
◇ A network with normal distribution or homogeneous degree
distribution (does not have distinct pattern).
200
201
Scale-free network
◇ In a scale-free network, few
central nodes control the follow
of data
◇ Facebook and Twitter Social
networks (few people with many
connections)
◇ Few websites having more
in-links (Google and Yahoo).
◇ Citation networks (few Banking activity network; nodes size represent
financial assets and links represent flow of capital.
scholars with many citations);
201
202
Small world network

◇ Small world networks have one or few dominant nodes and
many nodes with a relatively low level of degree
202
203
Small world network..

◇ Decentralized networks have the small-world property i.e., a
network in which most nodes are not neighbors, but most nodes
can be connected with a few steps
203
204
Common Social Media Network Types

◇ Friendship Networks
◇Follow-Following Networks
◇Fan Network
◇Group Network
◇Professional Networks
◇Content Networks
◇ Dating Networks
◇ Co-authorship Networks
◇Co-commenter Networks
◇Co-like
◇Co-occurrence Network
◇Geo Co-existence Network
◇Hyperlink Networks
204
205
Types of Networks
◇ The networks can be classified in a variety of ways:
￭ based on existence,
￭ based on direction of links,
￭ based on mode, and
￭ based on weights.
205
206
Types of Networks..
206
207
Types of Networks based on Direction

◇ Directed Networks: A network with directed links among
nodes is called a directed network
◇ Undirected Networks: In undirected networks, the links

among the nodes do not have any direction
207
208
Types of Networks based on Mode

◇ One-Mode Networks: A one-mode network is formed among
a single set of nodes of the same nature
◇ Two-Mode Networks : Two-mode networks (also known as
bipartite networks) are networks with two sets of nodes of
different classes
◇ Multimode Network: A multimode network is also possible
where multiple heterogeneous nodes are connected.
208
209
Types of Networks based on Weights

◇ Weighted Networks: In weighted networks, the links among
nodes bear certain weights to indicate the strength of association
among the nodes.
◇ Un-weighted Networks: In un-weighted networks, links
among nodes does not bear weights
209
210
Types of Networks based on Existence

◇ Implicit Networks:
Implicit networks do not exist by default (or are hidden) and need
to be intentionally constructed with the help of dedicated tools
and techniques
◇ Explicit Networks:
Explicit social media networks exist by default; in other words,
they are explicitly designed for social media users to be part of.
210
211
Network types Examples
211
212
Common Network Properties

◇ Node-Level Properties
• Degree Centrality
• Betweenness Centrality
• Eigenvector Centrality
• Structural Holes
212
213
Degree Centrality
◇ Degree centrality of a node measures the number of links a node has to the other
nodes in a network.
◇ In a Facebook network, for example, this will measure the number of friends that a
member has.
◇ In a Twitter network, it will equate to the number of followers or following a user has.
◇ In a directed network, the degree can be either in-degree (followers) or out-degree
(following).
213
214
Betweenness Centrality
◇ Betweenness Centrality of a node is related to the centrality (or position) it has in a
network. Nodes with high betweenness centrality can control the flow of information
between connected nodes due to their central position in the network.
214
215
Betweenness Centrality
Figure : ITM institutional collaboration network (node size represents

betweenness centrality) (Source: Khan G. F., Jacob W. (2016)
215
216
Eigenvector Centrality
◇ Eigenvector Centrality of a node measures the importance of a node
based on its connections with other important nodes in a network. It can
provide an understanding of a node’s networking ability relative to that of
others
E.g., Google search
engine use eigenvectors
to rank website based
on importance of in-links
216
217
Test Your Knowledge

Consider the following sample network:
 How many nodes are in the network?

 How many edges are in the network?
 Is this graph directed or undirected?
 Which node(s) have the highest
degree? What is the degree? Which
node(s) have the lowest degree?
What is the degree?
217
218
Common Network Properties

◇ Network-Level Properties
• Density
• Components
• Diameter
• Average Degree
218
219
Density
◇ The density of a network deals with a number of links in a
network. It is number of links present in a network divided by the
number of all possible links
◇ (for an undirected network, the number of all possible links
can be calculated as n (n – 1)/2); where n is the number of nodes
in a network).
◇ A fully connected network, in which each node is connected to
every other node, will have a density of 1.
219
220
Density
n (n – 1)/2); where n is the number of nodes in a network).
Possible links (complete graph) = ?

Density=?
220
221
Network Components
◇ Components of a network are the isolated sub-networks that connect within but are
disconnected between, sub-networks (Hanneman and Riddle 2005).
221
222
Diameter
◇ The diameter of a network is the largest of all the
calculated shortest path between any pair of nodes in a
network (Wasserman and Faust 1994).
◇ It can provide an idea of how long it would take for
some information/ideas/message to pass through the
network
222
Lambton College

Action Analytics and Search Engine Analytics
Session-9
224
Objectives
◇ To understand Action Analytics
◇ Describe different types of actions
◇ To understand Search Engine Analytics
◇ Search Engine KPIs
224
225
What is Action Analytics?

◇ Social media actions analytics deals with
extraction, analysis, and interpretation of the
insights contained in the actions performed by
social media users
◇ Can be used to measure popularity and
influence of a product, service, or idea over social
media.
225
226
What is Action Analytics?..

◇ Social media actions are of great value to social
media marketers because of their role in:
￮ increasing revenue
￮ brand value, and
￮ Loyalty
226
227
Common Social Media Actions

◇Like or “Like” buttons
◇Dislike
◇Share
◇Visitors, Visits, Revisits
◇View
◇ Clicks
◇Tagging
◇ Mentions
◇Hovering
◇Check-in
◇Pinning
◇Embeds
◇Endorsement
◇Uploading & Downloading
227
228
Common Social Media Actions..

◇ Clickstream analysis Analyzing aggregate data about which pages a website
visitor visits and in what order.
228
229
Common Social Media Actions..

◇ Clickstream analysis is used for:
o website activity
o website design analysis
o path optimization
o market research, and
o finding ways to improve visitor experience on the
website
229
230
Types of Search Engines
230
231
Types of Search Engines..

◇ Crawler-Based: Crawler-based search engines create their
databases or lists automatically, without any human intervention.
◇ Examples: Google.com and Bing.com
◇ They operate in three steps:
◇ web crawling,
◇ indexing, and
◇ Searching or ranking
231
232
Types of Search Engines..

◇ Directories: The listings in directories are manually compiled
and created by human editors.
◇ Examples: Yahoo Directory
◇ Meta-Search Engines: Metasearch engines compile and
display results from other search engines
◇ Examples: MetaCrawler, Mamma, and Dogpile
232
233
Search Engine Analytics (SEA)

◇ SEA can mean two things:
o Search Engine Optimization (SEO): Techniques used to
improve a website’s ranking in a search engine result page
o Search Engine Trend Analysis: deals with analyzing and
understanding the keywords people use in a search engine
233
234
Search Engine Optimization
(SEO)
◇ Search Engine Result Page (SERP) generally have two types
of results:
o Organic or free: Social Presence, Social Content,
Bookmarking, Keywords, Meta tags, etc.
o Nonorganic search results: Google AdWords, Bing Ads,
Facebook Ads, etc.
◇SEO increases Website Traffic
234
235
Search Engine Optimization
(SEO)..
235
236
Types of Organic SEO

◇ White hate SEO (ethical techniques)
◇ Black hate SEO (unethical techniques)
o Link spamming, fake reviews, fake key word,
etc.
236
237
Paid SEO Advantages

◇ Very fast: Get targeted visitors very fast
◇ Can yield highly profitable results
◇ Can be highly targeted
o Time of day
o Geographic area
o Keywords and phrases
o Immediate feedback
237
238
Paid SEO disadvantages

◇ Expensive
◇ Heavy competition: Competitive keywords
demand higher bids
◇ You pay regardless of any sales
◇ Traffic reduces when you stop paying
238
239
How to get listed on the top
(free)?
◇ Hundreds of “ranking factors”
◇ Mostly ranking happen based on relevance and
popularity:
 Keywords:
o Title page keywords/meta tags
o Keyword rich URL structure
Website structure and high-quality contents
 Social media presence/profiles: Fan pages, blogs,
Wikipedia articles, Twitter, etc.
239
240
Back links plus Quality

◇ Why is page B ranked higher, even it has only one in-
link from A?
240
241
Back links plus Quality

◇ Google’s PageRank algorithm predominantly relies on
the quality of incoming hyperlinks (or in-links) to rank
websites.
241
242
Website Keywords and Meta
tags
242
243
Search Trend Analytics

◇ It can help answer a variety of questions, including the
following:
1. How people search for your brand?
2. When does interest spike in your products or services?
3. Which keywords drive more traffic?
4. Which regions are interested in your brand?
5. What are trending topics over the Internet?
6. How are your competitors performing?
243
244
Search Engine KPI or Metrics
244
245
Conclusion
◇ To understand Action Analytics
◇ Describe different types of actions
◇ To understand Search Engine Analytics
◇ Search Engine KPIs
245
Session-10:
Capturing Value
with Location and
Mobile Analytics
Introduction
 LOCATION ANALYTICS, also known as spatial analysis or geo-

analytics is concerned with mapping, visualizing, and mining
the location of people, data, and other resources.
Sources of Location Data
 Postal Address
 Latitude and Longitude
 GPS-Based
 IP-Based
 Bluetooth/Sensors/Beacons/
Categories of Location Analytics
• Business Data Driven Location

Analytics
• Social Media Data Driven Location

Analytics
Business Data-Driven Location Analytics
Business data-driven location analytics deals with mapping, visualizing, and mining location data to reveal
patterns, trends, and relationships hidden in tabular business data (e.g., sales by regions, states, country).
Applications of Business Data-Driven Location Analytics:
1. Powerful Intelligence
2. Geo-Enrichment
3. Collaboration and Sharing
Powerful Intelligence
Using sophisticated mapping techniques, such as heat mapping,
data aggregation (e.g., aggregating data to regions), and color-
coded mapping, can generate powerful business intelligence.
Powerful Intelligence
 Using GPS trackers Warner enterprise

track more than 9000 tracks.
 -Truck Location: where the truck is

heading,
 -Cargo info: the driver’s name and the
hours of road time, and what type of
freight is being carried. This is crucial
for scheduling trucks and drivers.
 Mileage: Werner can now bill mileage

to customers more accurately
Geo-Enrichment
 Simple data maps can be enriched with customer data,

including demographic, consumer spending, lifestyle, and
locations.
2) Social Media Data-Driven Location

Analytics:
Social media data-driven analytics relies on social media location
data to mine and map location of social media users, content, and
data.
Two common sources:

• GPS
• IP
 Uses of Social Media-Based Location Analytics:

1. Recommendation Purposes
2. Customer Segmentation
3. Advertisement
4. Information Request
5. Alerts
6. Search and Rescue
7. Navigation
Location Analytics and Privacy
Concerns
Location Analytics raise serious privacy issues related

to collection, retention, use, and disclosure of
location information.
1. freedom of movement
2. freedom from being observed
What Is Mobile Analytics?
 Mobile analytics may refer to two

things,
 Mobile Web Analytics and
 Apps Analytics.

 Mobile Web Analytics

 Mobile web analytics is mostly focused
on characteristics, actions, and behaviors
of mobile website visitors; that is, the
visitors to the mobile version of your
company’s website.

 Apps Analytics
 Apps analytics deal with understanding and
analyzing mobile application users’ characteristics,
actions, and behaviors.
 Mobile apps are used:

 to drive sales
 improve brand affinity, and
 make purchases possible with a few swipes

Purpose of Apps Analytics
 Who are my users?

 Which countries are they from?
 What actions they are taking?
 How do my customers navigate  Which countries were top
in the app? performers in terms of in-app
 What are my in-app payments purchases?
and revenue?  Which application version leads to
more sales?
 How long do they stay on my
app?  How long do my users stay inside
my application?
 Which operator, operation
system, and devices they use?  How often do my users open my
app?
 What item is purchased the most?  How many users started a specific
number of sessions?
 How do my applications versions
compare to one another?
Types of Apps
 Apps can be classified mainly in two

ways:
• From the perspective of development
and deployment, and
• From the perspective of objectives.

Types of Apps
 Classification from the perspective of

development and deployment:
• Native Apps
• Web-based Apps, and
• Hybrid Apps.

Types of Apps
 Native Apps
 Native apps are specifically created
for and installed on mobile devices.
Pros: Cons:
• Available in apps • Are device specific.
stores. • Expensive to develop.
• Easy to monetize • Can only be accessed
• Fast through specific mobile
• Adjust well to their devices.
native platforms
Types of Apps
 Web-Based Apps
 Web-based apps look like natives apps, but in reality
they are websites optimized for mobile access.
Pros: Cons:
• Less costly to develop & • Not available in App stores
maintain.
• Slow
• Device independent
• Can be accessed from any • Hard to monetize them
mobile device.
• Can be accessed through
Internet browsers.
Types of Apps
 Hybrid Apps
 A hybrid app combines the functionalities of
both native and web-based apps.
 Available in app stores.

 Can be wrapping it into a native container.
 Can be used on any mobile device.
 Low development cost.

Types of Apps
 From the Perspective of Objectives

• Transaction-Oriented Apps
• Ads-Oriented Apps
• Information-Oriented Apps
• Networking-Oriented Apps
• Communication-Oriented Apps
• Entertainment-Oriented Apps
• Education-Oriented Apps
• Self-Improvement Apps

Characteristics of Mobile Apps
• Always On
• Moveable
• Location Awareness
• Focused
• Personalization
• Short-Term Use
• Easy to Use

Developing Your Own App
 Three Strategies
 Do-It-Yourself
 Outsource It
 Go Open Source

Chapter 11:
Capturing Value
with Hyperlinks
Analytics

Introduction
 HYPERLINKS are the pathways of social media traffic.
 Hyperlinks are references to web resources (such as a website,

document, and files) that users can access by clicking on it.
 Inter-linking
 Intra-linking
Introduction
 Hyperlinks serve as a symbolic meaning of validating or

endorsing the linked organization.
 Hyperlinks that exist between two organizational

websites reflect a
 Sense of validation
 Trust
 Bonding
 authority, and
 legitimacy
Types of Hyperlinks
 In-Links
 Out-Links
 Co-Links
Types of Hyperlinks
 In-Links
 In-links are the incoming
hyperlinks or links directed
toward a website or originated in
other websites.
Types of Hyperlinks
In-Links
 In-links are of great interest to social
markers, because they bring traffic
to a particular website.
Quality of in-links
Number of in-links
Types of Hyperlinks
 Out-Links
 Out-links are hyperlinks generated
out of a website.
Types of Hyperlinks
 Co-Links
 if two websites receive a link from a
third website, they are considered to
be connected indirectly.
 if two pages link to a third page, they are

also considered to be co-linking.
Hyperlink Analytics
 Hyperlink analytics deals with extracting,

analyzing, and interpreting hyperlinks (e.g.,
in-links, out-links, and co-links).
 The basic assumption:

 the number and quality of hyperlinks to a
website equates to its importance or value.
Types of Hyperlink Analytics
 Hyperlink Environment Analysis

 Co-Link Networks
 In-Links and Out-Links Networks
 Link Impact Analysis
 Social Media Hyperlink Analysis
 Hyperlink Environment Analysis
 Hyperlink environment analyses deal with a particular website or set of
websites.
 Two forms:
 co-links networks, and
 in-links and out-links networks.
Link Impact Analysis

Link impact analysis investigates
the web impact of a website
address (or URL) in terms of
citations or mentions it receives
over the web.
Social Media Hyperlink Analysis

Social media hyperlink analysis
deals with extraction and analysis
of hyperlinks embedded within
social media texts (.e.g., tweets
and comments).
Example: Social Media Hyperlink
Analysis
A comparative analysis
of out-links (embedded
in tweets) between
tweets of the Korean
and US governments by
Khan et al. (2014).
Source: Khan, G. F., H. Y. Yoon, et al. (2014).

"Social media communication strategies of
government agencies: Twitter use in Korea and
the USA." Asian Journal of Communication 24(1):
60-78.
Figure 2. Twitter networks for Korea (left) and the US Governments (right)
Hyperlink Analytics Tools
 Webometrics Analyst: Webometric Analyst is a web

impact analysis tool and can conduct variety of
analysis on social media platforms including hyperlink
network analysis and web mentions:
http://lexiurl.wlv.ac.uk/
 VOSON: VOSON (http://www.uberlink.com/) is a

hyperlink analytics tools for constructing and
analyzing hyperlink networks.
 Open Site Explorer: Open Site Explorer is a link

analysis tool to research and compare competitor
backlinks, identify top pages, view social activity data,
and analyze anchor text:
https://moz.com/researchtools/ose/
 Link Diagnosis: Link Diagnosis

(http://www.linkdiagnosis.com/) is a free online tool
for analyzing and diagnosing links.
 Advanced Link Manager: Advanced Link Manager

provides a variety of link analysis capabilities,
including the ability to track link-building progress
over time, domain quality analysis, backlinks
evolution, and website-crawling abilities:
http://www.advancedlinkmanager.com/
 Majestic: Majestic (https://majestic.com) provides a

variety of link analysis tools, including link explorer,
backlinks history, and link mapping tools.
 Backlink Watch: Backlink Watch

(http://backlinkwatch.com/) is a free tool for checking
the quality and quantity of in-links pointing to a
website.
Case Study: Hyperlinks and Viral
YouTube Videos
 The Problem
 The Solutions
 The Results
291
Tutorial: Hyperlinks Analytics with
VOSON
 To access the VOSON, you

must first create an account
by visiting this link:
http://www.uberlink.com and
click on the “create a new
account” option available at
the top of the page.
Figure 3. Hyperlink network constructed with VOSON

Copyright © 2015 Gohar F. Khan 292
Review Questions
 What are hyperlinks, and why they are important?

 Briefly discuss in-links, out-links, and co-links.
 What is hyperlink analytics and its underlying
assumptions?
 What is hyperlink environment analysis?
 What is link impact analysis?
 What is social media hyperlink analysis?
293
Chapter 12:
Capturing Value
with Multimedia
Analytics

Multimedia analytics
 Social media multimedia analytics is the art and

science of harnessing business value from video,
images, audio, and animations, and interactive
contents posted over social media outlets.

TYPES OF MULTIMEDIA ANALYTICS
 Three principal types of multimedia analytics are:

 Image Analytics
 Video Analytics
 Audio Analytics

Image analytics
 Image analytics deals with extracting business insights

(e.g., faces, logos, objects, scenes, and actions) from
images posted on social media.

Image analytics
Google Cloud Vision API
Figure 12.1. Google’s Cloud Vision API demo

General Uses of Image Analytics
 Label Detection
 Explicit Content Detection
 Logo Detection
 Optical Character Recognition
 Face Detection
 Image Attributes
 Web Detection

Business Uses of Image Analytics
 Brand Mentions
 Sentiment Analysis
 Measure Sponsorship ROI
 Find Visual Influencers
 Identify Moments of Consumption

Brand Mentions
Figure 12.2. Brand visual mention example

Sentiment Analysis
Figure 12.3. Visual sentiment example

Measure Sponsorship ROI
Figure 12.4. Sponsorship ROI example

Source: https://www.instagram.com/p/BPBTbYqALTB/
Find Visual Influencers
Figure 12.5. Visual influencer example

Source: https://www.instagram.com/p/BITU0zeB8ZQ/
Identify Moments of Consumption
Figure 12.6. Moments of consumption example

IMAGE ANALYTICS TOOLS
 CrimsonHexagon Image Analytics

 Simply Measured
 Iconosquare
 InstaFollow
 Image Memory value
 Pinterest Analytics
 Tailwind
 Curalate

VIDEO ANALYTICS
Video analytics can be divided into two categories:

 Video Content Analysis
 Video Stats Analytics

Video Content Analytics
 Video content analytics (VCA) is used to automatically

analyze contents of a video to detect and determine
temporal (time-bound) and spatial (space-bound)
events that are captured in a video.

Video Content Analytics..
Table 12.1. Application of video content analytics
Application Description
Recognition Video analytics is used for facial and object recognition (e.g., number plates) to recognize, and
therefore possibly identify persons or objects. For example, IBM video analytics tool allows users to
enroll facial images of objects and people of interest in a watch list and the system compares them with
faces captured by body cameras. High-quality matches are ranked for analyst review.
Detect changes in patterns From live-streaming fixed cameras, receive automatic alerts when movement of objects (people and
vehicles) is inconsistent with predefined patterns
Flame and smoke Internet-connected video surveillance cameras can be used to detect flame and smoke in 15–20 seconds
detection or even less because of the built-in smart microchip. The microchip processes are capable of analyzing
flame and smoke characteristics such as color chrominance, flickering ratio, shape, pattern and moving
direction.
Egomotion estimation Egomotion estimation is used to determine the location of a camera by analyzing its output signal.
Motion detection Motion detection is used to determine the presence of relevant motion in the observed scene.
Shape recognition Shape recognition is used to recognize shapes in the input video, for example circles or squares. This
functionality is typically used in more advanced functionalities such as object detection.
Style detection Style detection is used in settings where the video signal has been produced, for example, for television
broadcast. Style detection detects the style of the production process.
Tamper detection Tamper detection is used to determine whether the camera or output signal is tampered with.
Video tracking Video tracking is used to determine the location of persons or objects in the video signal, possibly with
regard to an external reference grid.
Video Stats Analytics
 Video stats analytics (VSA) is about monitoring and

analyzing the performance of videos posted on social
media channels, such as YouTube.

YouTube Analytics
 YouTube analytics provides very comprehensive data

for the YouTube channel owner including views, likes,
watch time, traffic sources, and demographics
reports.

YouTube Analytics
Figure 12.7. YouTube analytics

YouTube and Online Video
Advantages
 Low Production Costs

 Extensible
 Social Media
 YouTube
 Advertising Network
 Sponsorship Opportunities
 YouTube Production Studios
 YouTube Analytics
 Advocacy
 Search Engine Ranking

Issues with YouTube
 Stiff completion
 Low Attention Span
 Viral Sharing
 Big Brother
 Throttling and Network Congestion

Video Analytics Tools
 ChannelMeter
 Vidooly
 Quintly
 RankTrackr
 Socialbakers
 Ooyala

Audio Analytics
 Audio analytics deals with sound analysis and

recognition using advanced algorithms and artificial
intelligence.

Image Analytics
Figure 12.8. Image analysis

Analytics Types
Figure 12.9. Types of analytics and business value

Image Analytics Example
Figure 12.10. Descriptive image analytics example

Predictive Image Analytics
Figure 12.11. Predictive image analytics example

Prescriptive Image Analysis
Figure 12.12. Prescriptive image analytics example

Logo Recognition
Figure 12.13. Measure ROI with Logo recognition example

Source: http://www.twincities.com/2016/07/19/vikings-us-bank-stadium-ugly-on-the-outside-better-interior/
Image Analytics
Figure 12.14. understanding the full value of your sponsorships example

Moment Identification
Figure 12.15. Identify moments of consumption example

Data Driven Decisions using Image
Figure 12.16. Making data-driven decisions using image

Lambton College

Social Media, Security, Privacy, and Ethics
Session-12
328
Objectives
◇ Understanding Social Risks: Legal, Privacy, and
Security Risk
◇ Securing your social media
◇ Social Media Risks Management: Identity,
access, prioritize, and mitigate
328
329
Social Media Risks

◇ Risk: Risk is the possibility of losing something
of value such as, intellectual or physical capital
◇ Social media risk is the possibility of losing
something of value due to social media use.
◇ Privacy Risks
Legal Risks
Security Risks
Ethical Risks
Technological Risks
Social Risks
Economical Risks
329
330
Social Media Regal Risks

◇ Legal risk is the potential of losing something of value
(such as information, reputation, or goodwill) due to the
lack of knowledge of the way law and regulation apply
to a business.
◇ Legal risk may arise because of:
￭ Defamation: Libel (written form) and Slander(audio or
video)
￭ Spamming, copyright violence, etc.
330
331
Intellectual Property
◇ Intellectual property rights (IPRs) are the rights
granted to the creators of IP
￭ Copyright
￭ Trademarks
￭ Patents
◇ Materials could be ideas, documents, pictures, songs,
etc.
331
332
Trade Secrets
◇ Not generally known to public; it can be a formula, an algorithm, design,
recipe, process, method, etc.
◇ Can easily be leaked on social media, often mistakenly
◇ An employee leaking trade secrets through social media (or any
other medium) breaches:
• Duty of loyalty
• Breach of contractual confidentially
• Non-discloser agreement, etc.
332
333
Trade Secrets
◇ How to stop the breaching process?
￭ Updated social media use policies: (e.g., what
employees should not discuss about the company’s
plans on social media).
￭ Updated Non-discloser agreement: A legal
document prohibits material, knowledge, or information
exchange with third parties
￭ Training of the employees
333
334
Trademark
◇ Confusing customer about the trademark
￭ You maybe liable for creating a brand
handler/account similar to other brands that confuses
other social media users
￭ There may also be liability for the use of other
trademarks in Google Adwords, Keywords, and in
metatags
334
335
Spam
◇ Sending unwanted messages/posts
￭ A commercial message send without the
consent of the recipient is considered spam
◇ To comply:
￭ Obtain consent
￭ Provide opt-in and opt-out options
￭ Keep your messages header honest
335
336
Privacy
◇ “the right to be let alone.” (Warren and
Brandeis, 1890)
◇ Privacy in context of social media is the “ability
to decide what information one discloses or
withholds about oneself on social media, who has
access to such information, and for what purposes
one’s information may or may not be used.”
336
337
Accessing Social Media Data

◇ In what circumstances can we access social media
data?
◇ Only, if:
￭ The information is publicly available, OR
￭ consent is obtained, OR
￭ the information will be used for statistical or
research purposes, OR
￭ will not be used in a form in which the concerned
individual is identified
337
338
How can we stop third-party apps

from collecting our data?
◇ Don’t use them if you think the app is not that
important
◇ Disable/remove it from your profile
338
339
Securing your social media
platforms
◇ Use strong passwords
◇ Multi-factor authentication
◇ Trusted contact
◇ Disable or revoke third-party apps
◇ Login notification
◇ Review your login history
339
340
platforms
◇ Multi-factor authentication
￭ Extra layer of security
￭ If someone figures out a user password, they
will still not be able to access the account
unless they have physical access to a security
token such as 4 digit pin.
340
341
platforms
◇ Disable or revoke third-party apps
￭ Third-party apps are developed by other
companies but have access to Facebook via its
application programming interface (API).
￭ May handle your account information
insecurely
341
342
platforms
◇ Trusted contact:
￭ The trusted contact is an account recovery
feature provided by Facebook to help you
access your account securely through your
friends if you have trouble accessing your
account.
￭ Trusted contacts can send a code and URL
from Facebook to help you login
342
343
platforms
◇ Login notification:
￭ Enable your Facebook login notification so that you
can be notified through e-mail or text message when
your account is accessed.
◇ Review your login history:
￭ It is a good practice regularly to review your account
login history and location
343
344
Social Media Risk Management
Framework
◇ Social media risks management loop consists of four
iterative steps:
￭ Identify
￭ Access
￭ Mitigate, and
￭ Evaluate
344
345
Social Media Risk Management
Framework
• Periodic • Risk events

assessment and identification and
review understanding
Evaluate Identify
Mitigate Assess
• Risk mitigation • Risk impact and
planning and probability
strategy assessment
• Risk mitigation • Risk prioritization
implementation
345
346
Risk Assessment
◇ The risk assessment process determines the
likelihood of a social media risk event that could impact
the organization :
￭ Legally
￭ Economically
￭ Technically
￭ Politically, and
￭ Socially
346
347
Risk Assessment
◇ Risks are priorities and ranked based on probability
of occurrence and impact on an organization.
347
348
Risk Mitigation
◇ Typical risks mitigation strategies are:
￭ Risks management governance
￭ Training and awareness
￭ Social media policy
￭ Securing social media platform
348
349
Risk Management Governance

◇ New governance structures, roles, and policies should be created
within the business for properly managing social media risks. These
activities may involve
￭ Identifying and empowering a social media risk management
manager
￭ developing a business-wide risk-management strategy
￭ the identification of actions and steps needed to implement the
strategy, and
￭ determining the resources required to mitigate the risks (Garvey
2008).
349

Session 1

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Session 1

Uploaded by

Copyright:

Available Formats

Lambton College

School of Computer Studies

AML-3204 Social Media Analytics

Assignments (2 @ 10%) 20%

Final Term Project (1 @ 40%) 40%

Mark(%) Grade Mark(%) Grade

The Usage of Internet

What is Social Media?

Benefits of Social Media

◇Companies can reach a wider audience fast

Dos of Social Media

Types of Social Media

Issues of Social Media

Social Media Analytics

Social Media Analytics Tasks

Social Media Analytics Tasks..

Social Media Analytics Tasks..

In the next session we will study the following topic:

AML-3204 Social Media Analytics

Value of Social Media

The categories of value

Social Media Tangible V2F

More on Tangible V2F

Social Media Return on Investment

Social Media Return on Investment

Social Media Value Metrics

Social Media Engagement (SME)…

Social Media Influence (SMI)

Social Media Popularity (SMP)

In the next session we will study the following topic:

AML-3204 Social Media Analytics

What is Social Media Analytics?

What is Social Media Analytics?

Interest in Social Media Data

Purpose of Social Media Analytics

Types of Social Media Analytics

Case Study: Item Recommendation

Example: One-hot Encoding

In the next session we will study the following topic:

AML-3204 Social Media Analytics

Eight Layers of Social Media Analytics

Apps 8 Layers of Actions

Layer One: Text

Layer One: Text

Layer Two: Network

Twitter Network Example

Layer Three: Actions

Layer Four: Apps

Layer Five: Hyperlinks

Layer Six: Location

Layer Seven: Search Engines

Layer Eight: Multimedia

Social Media Monitoring

Social Media Monitoring tools

Social Media Listening

Social Media Analytics Cycle

Social Media Analytics Process

In the next session we will study the following topic:

AML-3204 Social Media Analytics

Utilization of Analysis Results

Utilization of Analysis Results..

Utilization of Analysis Results..