Random commentary about Machine Learning, BigData, Spark, Deep Learning, C++, STL, Boost, Perl, Python, Algorithms, Problem Solving and Web Search
Sunday, October 30, 2011
Can we use Facebook messages to predict how the stock market is going?
This would be an interesting Machine Learning problem on a massive dataset. Can we use Facebook's messages to predict how the stock and bond market will evolve? Here a couple of examples of the raw dataset.
Just curious what you meant by facebook messages ? Do you mean interaction between individual users ? If that's the case I am not sure how we might get our hands on that kind of data.
The links you provided, for the raw dataset, points to facebook pages of the respective websites. In which case there could possibly be some correlation between the comments on the respective facebook page to the performance of the stock itself, Is this what you might have referred to ?
If so, would they be temporally correlated to make it predictive or useful ?
Just curious what you meant by facebook messages ? Do you mean interaction between individual users ? If that's the case I am not sure how we might get our hands on that kind of data.
ReplyDeleteThe links you provided, for the raw dataset, points to facebook pages of the respective websites. In which case there could possibly be some correlation between the comments on the respective facebook page to the performance of the stock itself, Is this what you might have referred to ?
If so, would they be temporally correlated to make it predictive or useful ?
Just curious. :)
I meant the messages posted by the users. Just click the above links.
ReplyDelete