HISQA SYSTEMS AND METHODS FOR HUMAN INSPIRED SIMPLE QUESTION ANSWERING HISQA

摘要

The system and method for determining how to respond automatically to questions such as "Where does Harry Potter go to school?" Are discussed in this specification. A carefully constructed knowledge graph provides a wealth of factual resources, but there are still difficulties in answering fact-based questions in natural language, as questions can be raised in a wide variety of ways. Described herein are embodiments of a system and method for human-inspired simple question answering (HISQA), a deep neural network based method for automatically responding to a query using a knowledge graph. In the embodiment, first of all, inspected by the human action among these tasks, the correct object is searched through the object connection, and then an appropriate relation is searched for and the question is answered, and the two steps are all connected to the de- It is realized as a nerve embedding mechanism.