Golden-brand verbal skill recommendation method and device based on multivariate semantic fusion

Abstract

The invention provides a Jinjiang verbal skill recommendation method and device based on multivariate semantic fusion. The method comprises the steps that word segmentation and word vector initialization operation are conducted on historical dialogues, user questions of the current round and user attributes; based on a hierarchical semantic coding mechanism and a user attribute coding mechanism, dialogue semantic coding and user attribute semantic coding are carried out on the initialization operation result, and corresponding semantic representation is obtained; and fusing the coding results to obtain a fused semantic representation, and matching the fused semantic representation with the semantic representation of each verbal skill in a Jinjiang verbal skill library to obtain a recommendation result of the verbal skill. According to the method, a hierarchical semantic coding mechanism based on a bidirectional long-short-term memory network is designed, semantics of historical dialogues are effectively captured through word-level coding and sentence-level coding, and accuracy of verbal skill recommendation is improved by fusing the semantics into a subsequent verbal skill recommendation process. Based on the method, the invention further provides a gold brand verbal skill recommendation device based on multivariate semantic fusion.

Description

technical field [0001] The present invention relates to the technical fields of artificial intelligence and natural language processing, and in particular to a method and device for recommending gold-medal speech skills based on multivariate semantic fusion. Background technique [0002] At present, with the continuous development of natural language processing technology, various industries and enterprises have gradually begun to pay attention to the intrinsic value of dialogue data. Among them, telemarketing is an important scenario for dialogue data generation. A successful marketing dialogue should enable both parties to realize the value of telemarketing, and effectively improve the conversion rate and retention rate of customers. By observing related cases, it can be found that successful marketing dialogues are usually produced by experienced sales elites who have received professional training and rich practical experience. Great patience and mental quality. Consid...

AI Technical Summary

Problems solved by technology

[0010] The present invention provides a gold medal speech skill recommendation method and device based on multi-semantic fusion to solve the problem that existing speech skill recommendation methods ignore customer portraits and historical dialogue data Therefore, it limits the accuracy of the speech recommendation, and it is difficult to effectively realize the intelligent navigation of the speech and the promotion of orders.

Images