Man-machine conversation method, device, storage medium and computer program product

Abstract

The invention provides a man-machine conversation method, a man-machine conversation device, a storage medium and a computer program product. The method comprises the steps of determining a current conversation theme and current utterance information of a user; determining a current utterance representation vector of the user according to the current utterance information; in combination with thecurrent utterance information and the current utterance representation vector, performing graph reasoning calculation on the heterogeneous knowledge graph corresponding to the current dialogue theme,and selecting current knowledge corresponding to the current utterance information from the heterogeneous knowledge graph; acquiring current utterance information according to the current utterance information and current knowledge; generating a reply statement corresponding to the current statement, wherein the heterogeneous knowledge graph is created on the basis of structured knowledge and unstructured knowledge and can generate reply statements with rich contents. In addition, the accuracy of knowledge selection can be improved by adopting a graph reasoning algorithm, so that the knowledgeselection process has very good interpretability and generalization ability. Meanwhile, the dependence of the whole scheme on corpora with labels is reduced.

Description

technical field [0001] The present application relates to the field of computer technology, and in particular to a man-machine dialogue method, device, storage medium and computer program product. Background technique [0002] In the current dialogue system, for the dialogue topic, the corresponding background knowledge includes structured knowledge and unstructured knowledge, while the dialogue system only considers one of the knowledge, and vectorizes each knowledge in one of the knowledge to obtain each Knowledge representation of knowledge; obtain the discourse representation corresponding to the user's current discourse and the above information; according to the knowledge representation and discourse representation of each knowledge, combined with the neural network model, etc., select part of the knowledge from each knowledge to generate the reply sentence corresponding to the current discourse . [0003] However, in the above scheme, if only structured knowledge is ...

AI Technical Summary

Problems solved by technology

[0003] However, in the above scheme, if only structured knowledge is considered, because structured knowledge is generally an entity or a word, etc., it is impossible to generate rich reply sentences; if only unstructured sentences are considered, since unstructured sentences are generally sentences, Lack of structured organization, it is difficult to select appropriate knowledge, and the reply sentences generated based on multiple sentences are likely to be confused or inconsistent with the actual facts

In addition, the above scheme has the problems of low accuracy of knowledge selection and poor interpretability

Images