Symbiosis School for Online and Digital
Learning
SSODL
Title of the Topic – Social Media Analytics
Faculty Name – Ms. Radha Harilal Kollara
Symbiosis International University
17/12/2023 1
Social Media Analytics
Unstructured Data of Social Media and
Methods of Handling
Symbiosis International (Deemed University)
Session Objectives
• By the end of this session, you will be able to explain:
• What is unstructured social media data?
• What is unstructured data?
• Importance of social media data
• Types of unstructured social media data
• Methods of handling unstructured data
• Tools for unstructured data handling
• Best practices for unstructured data handling
Unstructured Social Media
Data
• Definition
• Unstructured social media data encompasses
information lacking a predefined data model, often
found in user-generated content across platforms
like Twitter, Facebook, and Instagram.
• Example: Analyzing unstructured data involves using
tools to make sense of user-generated content on
platforms like Twitter, Facebook, and Instagram.
What is Unstructured Data?
• Unstructured data lacks a formal structure and is
prevalent in the spontaneous, varied content found in
tweets, Facebook posts, and other user-generated
inputs.
• Example: Imagine a popular fast-food chain planning a
new menu launch. To understand customer preferences
and predict potential trends, the company decides to
analyze unstructured data from social media platforms
like Twitter and Facebook.
Why Social Media Data is Unstructured
• The freedom of expression on social media platforms
leads to diverse content styles, making it challenging to
organize and analyze.
• Example: Use techniques to organize and analyze the
spontaneous and diverse content found on social media
platforms.
Importance of Social Media Data
• Extracting Insights for Business Growth Social media
data analysis helps businesses glean valuable insights,
such as understanding the target audience and
identifying trends that drive informed decision-making.
• Example: Social media data analysis helps businesses
make data-driven decisions by examining user
engagement metrics, enabling the enhancement of
marketing strategies and overall business growth.
Types of Unstructured Social Media Data
• Textual Data (Posts, Comments): Unstructured textual
data includes tweets, comments, and posts, with
variations in language, slang, and abbreviations.
• Example: Apply sentiment analysis tools to understand
the positive or negative sentiment expressed in tweets
about a product launch.
Types of Unstructured Social Media Data
• Visual Data (Images, Videos): Visual content on social
media, from images to videos, presents challenges and
opportunities for analysis.
• Example: Utilize image and video analysis tools to
predict trends and measure the impact of visually
appealing content on user engagement.
• Task: Go to Myntra, or Meesho and check user updated
images. Analyze if there are common trends.
Types of Unstructured Social Media Data
• Audio Data (Voice Messages): Unstructured audio data
involves extracting information from voice messages,
such as customer feedback from recorded service
interactions.
• Example: Apply voice recognition tools to extract
valuable customer feedback from recorded service
interactions.
Methods of Handling Unstructured Data
• Natural Language Processing (NLP)
• NLP helps computers understand and work with
human language. It breaks down sentences into
smaller parts and identifies important things like
names and locations.
• Example: Use NLP to understand what customers are
saying in reviews. It's like teaching a computer to read
and analyze feedback to know if it's positive or
negative.
Methods of Handling Unstructured Data
• Image and Video Analysis
• This is about teaching computers to "see" and
understand pictures and videos. It helps recognize
objects and faces.
• Example:
• Imagine using this to see how people react in videos.
It's like a computer watching and understanding if
people are happy or sad.
Methods of Handling Unstructured Data
• Sentiment Analysis
• Sentiment analysis is like a computer figuring out if a
piece of text (like a review or comment) is expressing
a positive, negative, or neutral feeling.
• Example:
• Apply sentiment analysis to know how customers
feel about a product. It's like the computer reading
comments and telling you if people like it or not.
Methods of Handling Unstructured Data
• Topic Modeling:
• Topic modeling helps identify topics present in a
collection of texts. It's useful for understanding the
main themes or subjects discussed in unstructured
text data.
• Example:
• Apply topic modeling to news articles to
automatically identify and categorize articles based
on their main topics.
Methods of Handling Unstructured Data
• Pattern Recognition:
• Pattern recognition involves identifying patterns or
trends within data, which is useful for extracting
meaningful insights.
• Example:
• Use pattern recognition to identify trends in
customer behavior based on historical data, helping
in decision-making.
Tools for Unstructured Data Handling
• Data Analysis Tools
• Python libraries (NLTK, spaCy) and social media
analytics platforms (Hootsuite, Sprout Social) are
crucial tools for handling unstructured data.
• Popular Social Media Analytics Platforms
• Example: Hootsuite for scheduling and analyzing social
media posts and is effective in handling unstructured
data.
• Task: Open Hootsuite and explore
Best Practices for Handling Social Media
Data
• Data Privacy Considerations:
• Adhering to data protection regulations and
implementing anonymization techniques
• Example: Ensure compliance with data protection
regulations, such as GDPR (General Data Protection
Regulation), by implementing anonymization
techniques and safeguarding user data during handling
processes.
Best Practices for Handling Social Media
Data
• Continuous Monitoring and Updates:
• Real-time analytics are crucial for staying informed
about evolving user behaviors.
• Example:
• An e-commerce platform that has just launched a flash
sale for a limited-time discount on selected items.
Suddenly, there is a spike in user engagement on a
specific product category that wasn't anticipated.
Case Study - Company: Spotify
• Data Source: Unstructured social media data from
Twitter, Instagram, and Facebook, including user posts,
comments, shares, and reactions related to music
preferences, artists, genres, and listening habits.
Spotify - Best Practices Employed
[Link] Collection:
[Link]-time monitoring: Spotify utilizes social listening
tools to capture real-time conversations and trends
around music on various platforms.
[Link] integration: They integrate with social media
APIs to access a wider range of user data and
interactions.
[Link] reach: Data collection encompasses diverse
languages and regions to cater to Spotify's global
audience.
Spotify - Best Practices Employed
[Link] Cleaning and Preprocessing:
[Link] removal: Irrelevant information like
spam, bots, and off-topic content are filtered out.
[Link] analysis: Techniques are used to identify
positive, negative, or neutral sentiment towards
artists, genres, and overall music experience.
[Link] recognition: People, songs, albums, and other
relevant entities are identified and categorized for
further analysis.
Spotify - Best Practices Employed
[Link] Analysis and Insights:
[Link] identification: Spotify analyzes the data to
discover trending artists, genres, and topics in
different regions and demographics.
[Link] modeling: Machine learning algorithms
are used to predict user preferences based on their
social media activity and listening history.
Spotify - Best Practices Employed
[Link]: Insights gained from social media
data are used to personalize user
recommendations, discover playlists, and curate music
content tailored to individual tastes.
Spotify - Best Practices Employed
[Link] and Optimization:
[Link] creation: Music editors and curators
leverage social media trends to create genre-specific
playlists, artist spotlights, and other engaging
content.
[Link] and outreach: Targeted advertising
campaigns are launched based on user preferences
derived from social media data.
Spotify - Outcomes and Benefits:
• Increased user engagement: Personalized
recommendations and relevant content keep users
engaged and coming back for more.
• Improved user experience: Social media data helps
Spotify understand user preferences better, leading to a
more enjoyable and satisfying listening experience.
Spotify - Outcomes and Benefits:
• Discovery of new talent and trends: By staying ahead
of the curve with social media insights, Spotify can
promote rising artists and cater to evolving musical
tastes.
• Enhanced brand reputation: Proactive engagement
with users on social media fosters a positive brand
image and strengthens customer loyalty.
Key Takeaways
• Overall, Spotify's case study demonstrates how
effectively applying best practices to unstructured
social media data can:
• transform user experience
• drive engagement,
• and fuel their continued success in the ever-evolving
music streaming landscape.
Session Outcomes
In this session, you learned about:
What is unstructured social media data
Importance of unstructured social media data
Types of unstructured social media data
Methods of Handling unstructured social media data
Case Study
Thank You