As we defined earlier, the RB system is developed with the help of human domain experts. Let's take some examples in this section which can help to prove our rule of thumb:
- Say, we want to build the machine translation system from English to available Indian corpora and they are too small. The translation system should be accurate enough in order to develop it. We need human experts who know English as well as Gujarati. We don't want to address all the different levels of translation at a time, so we need to cover small chunks of the problem first and then on top of the developed prototype, we will build other chunks. So, here also, I would like to choose the RB ...