Applying Word Embeddings to Leverage Knowledge Available in One Language in Order to Solve a Practical Text Classification Problem in Another Language.