2. Agenda
⢠Social Media: An INT perspective
⢠Common Analytic Pitfalls
⢠An Analytic Framework
⢠Case Study: Brand Management
â Problem Definition
â Source Selection
â Data Capture
â Data Reporting
â Data Analysis
⢠Ways Forward, Future Analysis
⢠Questions?
3. Intelligence
⢠Intelligence is information that has been
transformed to meet an operational need
Operational Lens
Data Intelligence
5. Social Media: The INT Perspective
Social Media gets the best
and worst of three disciplines:
HUMINT
â HUMINT
⢠Pros: Reveals intentions
⢠Cons: Can be unreliable
â OSINT
⢠Pros: Fast, Accessible
OSINT SIGINT ⢠Cons: Noise
â SIGINT
⢠Pros: Network, High Volume
⢠Cons: Noise
6. Social Media Analysis Goals
⢠Need to have an end-goal with value to the
organization (operational lens)
⢠Need to ensure cyclical feedback occurs from
collection, processing, analysis, and
consumption
⢠Need to make sure that a particular network is
the right source for the task
7. Common Misconceptions
⢠Social media is not a panacea
â Not everyone uses social media
â Users of social media use it unevenly
â User behavior changes based on situations
⢠Just because people can talk about anything
does not mean they talk about everything all the
time.
8. Common Pitfalls
⢠The important thing is often not what people are
saying⌠but why they are saying it.
⢠Reporting tools rarely help dig into the why.
⢠Many common tools, reports, and metrics are
actually misleading:
â Word clouds atomize message context
â Sentiment metrics are often highly inaccurate
â Information in aggregate hides more than it reveals
9.
10.
11. Dangers of Disintegration
Source: Matthew Auer, Policy Studies Journal,
Volume 39, Issue 4, pages 709â736, Nov 2011
12. Analytic Framework
⢠Data Capture (DC)
⢠Data Reporting (DR)
⢠Data Analysis (DA)
â 1. What to measure
â 2. What the data is saying
â 3. What should be done based on the data
Source: Avinash Kaushik, Occamâs Razor Blog
http://www.kaushik.net/avinash/web-analytics-consulting-
framework-smarter-decisions/
14. Choosing a Platform
⢠Social media is still new, evolving; and so
is how we use it.
â Static approaches to social media are flawed
from the outset
â No one metric or set of metrics will always let
you know what is happening
⢠Need an adaptive platform to facilitate
data capture, reporting, and analysis
15. Case Study: Brand Management
⢠Industry: Gaming
â Experiencing 10% growth annually
â Overall revenue expected to exceed $80
billion by 2014
⢠In May, Zenimax Online Studios
announced Elder Scrolls Online
â Elder Scrolls V: Skyrim 2nd largest game of
2011
16. Problem Definition
⢠As a brand manager, how can I use social
media to track and understand public
attitudes toward my product?
⢠Challenge is getting relevant information
â Query too large = false positives
â Query too small = miss potential information
17. Source: Twitter
⢠Twitter has some of the best
analytic potential
â High volume traffic
â High volume user-base
â Open API
⢠Not without limitations:
â 140 characters
â Limited historical / lookback
18. Platform: Infinit.e
Infinit.e is a
scalable
framework for Visualizing
Analyzing
Retrieving
Enriching
Storing
Collecting
Unstructured documents
&
Structured records
19. Platform: Infinit.e
⢠Infinit.e supports the extraction of entities
and creation of associations using a
combination of built in enrichment libraries
and 3rd party NLP APIs.
20. Data Capture â Initial Query
⢠Twitter search for âElder Scrolls Onlineâ
â Simplest possible way to access information
â RSS feed for 10 days (Jun 27 â July 6 2012)
22. Data Capture â Entity Map
Hashtag TwitterHandle URL
Who
TwitterHandle
What
Hashtags, Keywords,
URLs
When
Time, Date
Unstructured Keywords Where
Time / Date Stamp Geo (if Available)
23. Data Reporting
⢠Used Infinit.eâs Flash U/I Widget Framework
â Document Browser (Individual Tweets)
â Entity Significance (Top Entities)
â Sentiment (Top Entities w/ Sentiment)
â Query Metrics (Breakdowns of Query Results)
⢠Framework allows for additional
visualizations to be constructed as needed
⢠Export options also available for manual
review (e.g. graphml, excel, pdf)
27. Data Analysis
⢠Analysis needs to be rooted in the
operational need:
âHow can I use social media to track and
understand public attitudes toward my
productâ
⢠Emphasis on hypothesis generation,
testing, and experimentation
28. Data Analysis -> Capture
⢠Hash tags from an initial subset of Tweets
fed back into the initial query
Initial
Expanded Query
Query
Results
Results
Twitter
29. Data Analysis - Hashtags
⢠Top hashtags were
almost all generic /
more abstract
â Undermines tracking and
understanding
â Top hashtags tied to
franchise, not to the
game
30. Data Analysis - Sentiment
⢠Converted URLs into derivative sources
⢠35% additional sources
⢠Larger text sources offer potential value with
sentiment analysis that tweets alone cannot offer
31. Data Analysis - Sentiment
⢠Top negative and positive scores provided
glimpses into aggregate attitudes
⢠Provide starting points for additional analysis
32. Data Analysis - Recommendations
⢠Actionable recommendations allow
decision makers to make changes
33. Future Data Analysis
⢠Initial conclusions should be starting points
for new analysis
⢠Broad entity capture allows for:
â Key influencer identification
â Clustering of tweets for segmentation
â Map / Reduce for aggregate functions
35. Expandable Model
⢠Identify key influencers on specific topics
⢠Look at relationships between websites /
blogs and Twitter use (cross-network
analysis)
36. Counting and Summing
⢠âTraditionalâ business intelligence analytics
problems solved using aggregate functions:
â Sum
â Count
â Average
â Min
â Max
â Etc.
37. Clustering - Topic
⢠Topic Extraction
â Key words -> Categories
â Categories -> Related Categories
Keyword Topic Key Value
graphics graphics graphics gameplay.pdf
screenshots graphics story gameplay.pdf
resolution graphics company corporate.txt
quests story ⌠âŚ
zenimax company ⌠âŚ
⌠âŚ
39. Take-Aways
⢠All data providers can and do change their
formats; users flock to and abandon
platforms â what works today may not
work tomorrow.
⢠Whatever platform you choose to do
analysis, make sure itâs open and
adaptable or your investment may
degrade over time.
40. Take Aways (Things to Avoid)
⢠Data puking (less is more)
⢠Metrics that cannot be tied to actions
⢠Visualizations / reports that remove
context
⢠Taking dashboards at face value
41. Take Aways (Things to Do)
⢠Segment data rather than work in aggregate
⢠Look for the why behind the message
⢠Always return to the source material
⢠Explore alternative explanations
⢠Always consider the ultimate goal
42. Thank You!
Andrew Strite
www.ikanow.com
astrite@ikanow.com
github.com/ikanow/Infinit.e
Hinweis der Redaktion
Given my background, I come at the social media problem from an intelligence analysis perspective. This comes with a certain set of vocabulary and paradigms, but I believe they are useful for understanding how to frame out an effective analytic framework.