This example filters for positive and negative sentiment in the text. This also limits results to English and samples 30% of the total volume of social interactions.