NLP innovation: Using Implicit/Explicit Semantics to improve Machine Reading Comprehension and Response Selection
, NTT Resonant
, NTT Resonant
We'll talk about these two natural language processing topics: 1. Machine reading comprehension methods have gained much attention. They, however, do not investigate the implicit semantics among multiple passages in the answer generation process, even though topics correlated among the passages may be answer candidates. Our method determines which tokens in the passage are matched to other passages assigned to the same question and investigates the topics in which they are matched. Evaluations using MS-MARCO indicate that our method generates promising results. 2. To naturally understand the dialogues, natural language understanding systems must handle multi-turn utterances of users. Our method improves Bert to handle multiple utterances and also utilizes explicit semantics behind the word (i.e., WordNet synsets) into the Transformer's encoder. Thus, it can analyze the common sense shared among users in the conversations. Our evaluation using the Reddit dataset found that our model achieved high response selection accuracies.