Random Sampling and Inductive Ability Evaluation of Word Embedding in Medical Literature

In the NLP (natural language learning) applied to medical literature semantic analysis, we proposed a method to analysis different data base training result in order to show that RNN with rough set training method can learn some relations between medical nouns. In our previous work, we trained a RNN model by rough set data preprocessing in medical literature analysis. So, we improved the word embedding method by using random sampling and inductive ability evaluation.