Jillian M. Ketterer [email_address] October 13, 2009 What can we learn from sentiment analysis?
What is sentiment analysis? “… translating the vagaries of human emotion into hard data” [primarily in the online world]  Mining the web for feelings, not facts (NYT) “… us[ing] automated tools to discern, extract and process attitudinal information found in text…” Sentiment Analysis: Opportunities and Challenges (Grimes, 2008) “… an attempt to automatically process and possibly learn from the universe of people’s online chatter”   (Me, 2009)
 
In the virtual (online) world, organizations and individuals ...have access to the universe of shared opinions Reviews, comments, tweets, blog posts, tags … as well as behavior(s) demonstrated by users Clicks, purchases, browsing habits, social networking decisions (i.e. Share or not share? Follow or not follow?) … over time, and often in real time, Archive.org, Google/Twitter trends, RSS … in a readily analyzable format.
Academic Research (e.g., not sure of purpose)
Personal Use
Market Research/Business Intelligence Marketing Customer satisfaction Prediction markets Economics Computational Linguistics Semantic web (incorporating “feelings” into “meaning” – this could be big)
Powering Search Engines
Real-Time, Longitudinal Data Online Theoretically, this world exists.
Questions to Ponder What is the relationship between what you say online, and what you do (online and offline)? What can  be learned from data about “feelings”? How do sentiments differ across domains? Potential applications? Scouting? Reputation Management? Evaluation/ Assessment?
(Pssst Google is getting sentimental) BANG!  Wow ‘em with a headline

What can we learn from sentiment analysis?

  • 1.
    Jillian M. Ketterer[email_address] October 13, 2009 What can we learn from sentiment analysis?
  • 2.
    What is sentimentanalysis? “… translating the vagaries of human emotion into hard data” [primarily in the online world] Mining the web for feelings, not facts (NYT) “… us[ing] automated tools to discern, extract and process attitudinal information found in text…” Sentiment Analysis: Opportunities and Challenges (Grimes, 2008) “… an attempt to automatically process and possibly learn from the universe of people’s online chatter” (Me, 2009)
  • 3.
  • 4.
    In the virtual(online) world, organizations and individuals ...have access to the universe of shared opinions Reviews, comments, tweets, blog posts, tags … as well as behavior(s) demonstrated by users Clicks, purchases, browsing habits, social networking decisions (i.e. Share or not share? Follow or not follow?) … over time, and often in real time, Archive.org, Google/Twitter trends, RSS … in a readily analyzable format.
  • 5.
    Academic Research (e.g.,not sure of purpose)
  • 6.
  • 7.
    Market Research/Business IntelligenceMarketing Customer satisfaction Prediction markets Economics Computational Linguistics Semantic web (incorporating “feelings” into “meaning” – this could be big)
  • 8.
  • 9.
    Real-Time, Longitudinal DataOnline Theoretically, this world exists.
  • 10.
    Questions to PonderWhat is the relationship between what you say online, and what you do (online and offline)? What can be learned from data about “feelings”? How do sentiments differ across domains? Potential applications? Scouting? Reputation Management? Evaluation/ Assessment?
  • 11.
    (Pssst Google isgetting sentimental) BANG! Wow ‘em with a headline