i mean i have twitter data-set of " food price crises " which i've extracted using Apache Flume.
as a part of my MTech thesis , i m trying to identify Uncertainty in this dataset and sentiment analysis. but for this task, i need annotated corpus which is not publicly available.
i am trying to implement Fuzzy rule based classification system ( Machine Learning Approach ) for uncertainty identification.
now i m in trouble because of this corpus .
can u help me in this task , how can i proceed for uncertainty identification ?
I did not get the relationship between Uncertainty and Sentiment analysis? I guess those are different research areas. If you have any question about sentiment analysis, I may be able to help you.
You want (a) to find an already annotated corpus, or you want (b) your corpus to be annotated?
For (a), there are some available http://www.sananalytics.com/lab/twitter-sentiment/ ... For (b) either you use some already existing approach like the one pointed out by Erik (you will need to filter out mistakes, probably), or you involve human annotators (always challenging and time consuming but also more accurate if task is well defined)
Firstly, you need to have an expression for Uncertainty from Economics point of view. Economic Uncertainty is directly proportional to monetary inflation, irrespective of however it is caused. Inflation leads to lesser amount of affordability of Risk which means smaller gains and hence consequently in-competitive economy and also consequently germination of economic crises if this aspect affects asymmetrically any specific root aspect that has the tendency to govern most other economic parameters. Basically, you can model such uncertainities by observing your food price co-ordinate system relatively from a reference fame that takes note of the fluctuation of the inflation.