Question and answer method and device based on artificial intelligence, computer equipment and storage medium

Abstract

The invention discloses a question and answer method and device based on artificial intelligence, computer equipment and a storage medium, and the method comprises the steps: carrying out the languagemodel training based on first training data, carrying out the NER model training based on second training data and a trained language model, and carrying out the relation matching model training; identifying an entity in a to-be-processed statement based on the trained NER model, and obtaining a relationship corresponding to the to-be-processed statement based on the trained relationship matchingmodel; and determining and outputting an answer corresponding to the to-be-processed statement according to the relationship corresponding to the to-be-processed statement and the entity in the to-be-processed statement. Based on language model transfer learning and graph transfer learning technologies, a training method commonly used by a language model is improved, higher accuracy can be achieved through a small amount of manual marking data, and the method is more suitable for constructing a knowledge graph question-answering system.

Description

technical field [0001] The present invention relates to the technical field of knowledge graphs, in particular to an artificial intelligence-based question answering method, device, computer equipment, and storage medium. Background technique [0002] Knowledge map, also known as scientific knowledge map, is called knowledge domain visualization or knowledge domain mapping map in the library and information industry. It is a series of different graphics that show the process and structural relationship of knowledge development. Because it can provide high-quality structured data, knowledge graphs and question answering systems based on knowledge graphs are used in more and more fields, such as automatic question answering, search engines, and information extraction. A typical knowledge graph is usually expressed in the form of a triplet of head entity, relationship, and tail entity (eg Yao Ming, nationality, China). The expression in this example reflects the fact that Yao M...

AI Technical Summary

Problems solved by technology

[0003] However, the current knowledge graph question answering technology is still in the stage of exploration and research and development, and most of the achievements and progress are still based on academic papers. The specific plan is: according to the questions put forward by users, the corresponding papers or websites can be obtained through keyword searches in the database. Literature, users click on the specific content of the paper to find the content they need, which will lead to poor processing efficiency of the user's questions and cannot meet the user's requirements

[0004] Knowledge graph question answering system, regardless of open domain or vertical domain, accuracy is the main factor limiting its wide application, and the main factor of insufficient accuracy is that the amount of labeled data is too small

Images