Neurolingo is in the business of creating actionable information from unstructured text.
In the case of our SentiBet demonstration, we are focused on creating a forecast about the outcome of sporting events by determining the collective opinions/sentiment of sports fans from Twitter.
Here are the steps:
- taking unstructured text from the Twitter stream
- find Tweets related to specific sports and specific contests/games
- determine the forecast and type of forecast for each Tweet
- produce the Sentiment Based Forecast (SBF) for each event.
So how do we know what each Tweet is forecasting?
- We are looking for 3 types of Tweets:
- We have an algorithm that weights the value of each type of Tweet and then creates the SBF for each game.
- Our algorithm is constantly being refined
We also give the ability for customers to develop their own algorithm by clicking on the SBF Preferences section of each game.
Subscribed users can set up their own default SBF preferences through the myAccount page.