In this chapter, I described basic NLP concepts and demonstrated a few basic techniques. I hoped to demonstrate that pretty complex NLP concepts could be expressed and tested in a few lines of Scala code. This is definitely just the tip of the iceberg as a lot of NLP techniques are being developed now, including the ones based on in-CPU parallelization as part of GPUs. (refer to, for example, Puck at https://github.com/dlwh/puck). I also gave a flavor of major Spark MLlib NLP implementations.

In the next chapter, which will be the final chapter of this book, I'll cover systems and model monitoring.