An Evaluation Benchmark for Testing the Word Sense Disambiguation Capabilities of Machine Translation Systems