Why empty RDD is always sent to Spark when integrating Spark Streaming and Twitter API?

Question

I am new to Spark Streaming and I am doing a small personal project to study the technology. I want to obtain realtime tweets using Twitter API and then transform the stream dat