Beyond Brands: Listening Big for the Issues
Big Listening notes
What rachel is looking for: "like SETI project for the internet"
Rachael is working on a project "to make the Ocean famous on the internet"
Thinking about "Conversation Island": intersection of topic, audience, platform and period of time Some are long standing (topic on Redit) Others bubble up and die off. (blog post)
Racheel currently using Radian6 Corporate traffic/sentiment analysis (reads twitters, bogs, Clout, etc). Very expensive. Big fancy tool. $30,000 to set up, $10,000 per month. But very complicated pricing model, hard to estimate.
Doing social media monitoring directly is feasible, but time expensive.
What tools to people use? Google key words, blog search. Twitter. (hashtags: often disappointing 'cause they are not always actually conversations, follow people) Side note: (have to think about sub-keywords for topic. i.e. "Ocean" includes "starfish", "marine areas", etc) Google Alerts for news topics. (can get either as RSS feed or email) Online searches. Track news articles in spreadsheet, list keywords
What do we do with the data once it is collected? Listening allows individual to gain insight over time. Rachael: have to actually engage in listing to the real conversation, not just use high level tools.
Application: track everyone who comes to dev summit, see what topic trends they post and communication about. people used to use #nptech, and then Beth would summerize discussion topics for each week as a blog post.
Rachael demos a paper-based diagram from tech-soup. Nodes standing for various actors.
- Built google doc of feeds (with yahoo pipes to do language translations) its public. Url?
- Piped feeds into "listining dashboard" which was a NetVibes account
- Human would look at dashboard each day, and cut and past into google docs. i.e. Look for mentions of major donor.
Similar to Asperations listing dashboard, but this for tracking what what other people say on the web (not just post on channels that you control)
What is being done with NLP and Sentiment Analysis? What tools can we use to sift out relevant information?