Implicit Sentiment Mining in Twitter Streams

For our November Meetup, we're very happy to have Maksim (Max) Tsvetovat from local analytics consulting firm Deepmile Networks, talking about extracting sentiment from Twitter data. Although the idea of using billions of tweets to learn about opinions is appealing, getting it to work in a compelling and valuable manner has been fraught with difficulty. Max will bring us up to speed, and discuss a method that works well for certain domains.

Notes: We're back at Google for this event! And we'll be continuing our experiment with informal pre-event themed networking -- please come early to meet and chat with people interested in Natural Language Processing!

Agenda:

  • 6:30pm -- Networking and Refreshments (Discussion theme: NLP)
  • 7:00pm -- Introduction
  • 7:15pm -- Max's presentation and Q&A
  • 8:30pm -- Post presentation conversations
  • 8:45pm -- Adjourn for Data Drinks (location TBA)

Abstract:

In this talk, I will describe a new method for estimating sentiment in online speech. This method does not rely on pre-defined lists of "good" or "bad" words -- but, rather, measures affinity toward a subject, brand, politician, etc. by locating and measuring psycholinguistic similarities between speakers and producing aggregate sentiment statistics. This method is ideally suited to understanding sentiment toward politicians, journalists,  advertisers -- anyone that produces large amounts of direct speech. While this limits the domains in which this method is applicable, its accuracy
increases significantly.

Bio:

Max is the Chief Technology Officer at DeepMile. He has a PhD from Carnegie Mellon University and is currently a Research Assistant Professor at George Mason University where he teaches Social Network Analysis. He is widely published in computer science, organizational theory and social network journals, and is a regular presenter at industry conferences. To learn more about Max and his research, you can explore his website -- www.tsvetovat.org. You should also buy his book, Social Network Analysis for Startups.

Join or login to comment.

  • Janet D.

    Max Tsetovat will be teaching an online course, "Social Network Analysis Using Python" at Statistics.com that starts 9/20 - 10/18/13. More details:
    http://www.statistics.com/sna-python/

    June 10, 2013

  • Andy

    Similar topic on Twitter analysis, UC Berkeley course
    http://blogs.ischool.berkeley.edu/i290-abdt-s12/about/

    December 16, 2012

  • Harlan H.

    The audio for this event is now available on the files page: http://www.meetup.com/Data-Science-DC/files/ Also, if anyone is interested in writing up an event summary/review for the Data Community DC blog, please contact me!

    November 24, 2012

    • John Mark M.

      Hey, thanks so much for recording this. It is much appreciated.

      November 24, 2012

  • Maksim T.

    Just checked in a URL bug fix and code cleanup by @JackieKazil -- check GitHub for updates. The URL fix changes the results slightly, with more slant toward Israel and less toward Hamas. However, FOX News is still the most radicalized network!

    November 20, 2012

    • A former member
      A former member

      Max, thank you for great presentation and answering the questions.

      November 20, 2012

    • freddie s.

      Or, FOX News has the most unbiased / transparent reporting of what both sides are saying, thus resulting in a nearly 50/50 split. As was pointed out at the meetup, mere repetition of bi-grams may reflect quotes that are neutral or ironic as well as empathetic "mirroring."

      November 21, 2012

  • Phil K.

    Maksim Tsvetovat gave a very nice presentation. Clearly, algorithms to understand sentiment in natural language are progressing very quickly.

    November 20, 2012

  • freddie s.

    Entertainment value five stars, but information value only two. 3 attended from our team and can't agree if "mirroring" is an approximation, substitute or supplement to sentiment analysis. Implied interpretation of "media bias" analysis fell apart under questioning - clearly lots of feature refinement too be done. Overall wish Dr. T had stuck with the more conventional and well analyzed election example - attempt to be current was admirable but produced much more noise than signal.

    November 20, 2012

  • Maksim T.

    I added the slides to the repo. For some reason they didn't go through the first time. Also, see Slideshare -- http://www.slideshare.net/jazzflute/implicit-sentiment-mining-in-twitter-streams

    1 · November 20, 2012

  • Maksim T.

    Guys -- slides and code from the presentation are posted on GitHub: https://github.com/maksim2042/DC_DataScience_Meetup

    Enjoy!

    2 · November 20, 2012

    • Abhijit

      Same here. Couldn't find the slides

      November 20, 2012

  • Bill M.

    Great presentation -- very thought provoking.

    November 20, 2012

  • A former member
    A former member

    The meetup was great. Enjoyed the subject and the speaker. Need a microphone, though. People were asking questions and many of them were not audible to the rest of the attendees. Also the speaker need a lavalier.

    November 20, 2012

  • A former member
    A former member

    good presentation

    November 20, 2012

  • James S.

    this was really well run

    November 20, 2012

  • Ross M.

    Relevant and timely technical topic. Great tools and approach. Would like to hear more and a lot more in depth about this topic. Audience engaged early with good questions. This interesting speaker's excellent topic would have benefitted from a faster, denser and more linear presentation style.

    November 19, 2012

  • Loren

    Will the presentation and post presentation conversations be available for live of later viewing?

    November 19, 2012

    • Jackie K.

      Max is usually pretty good as making slides available. If not, you can ping him on twitter... @maksim2042

      November 19, 2012

  • David H.

    There's a garage on 11th Street, next to the Methodist church -- look for the blue/white signs. But really, with the downtown traffic you're better off taking Metro.

    November 19, 2012

  • Harry D.

    Any suggestion on where to park?

    November 19, 2012

    • David H.

      There are a couple of garages on 11th street that offer after-hours rates.

      November 19, 2012

  • John Mark M.

    What an amazing lineup! I am unable to make it to this meetup but would love to hear the material. Would anyone be willing to record it and post it for the rest of us that are unable ot attend?

    November 19, 2012

  • Amrinder A.

    Diakopolous and Shamma had given an approach after using the 2008 presidential debates and the related Twitter feeds. Looking forward to see Max's approach to sentiment mining.

    November 13, 2012

  • Jim B.

    Looks like a great gathering and I am sorry that a previous commitment keeps me from attending.

    November 13, 2012

170 went

Our Sponsors

People in this
Meetup are also in:

Create your own Meetup Group

Get started Learn more
Katie

I'm surpris ed by the level of growth I've seen since becoming an organizer, it's given me more confidence in my abilities.

Katie, started NYC ICO

Sign up

Meetup members, Log in

By clicking "Sign up" or "Sign up using Facebook", you confirm that you accept our Terms of Service & Privacy Policy