An intelligent voice dialogue data operation analysis method

Abstract

The invention discloses an intelligent voice dialogue data operation analysis method. The method comprises the steps that an intelligent voice dialogue system sends intelligent voice dialogue data ofeach round to a message queue; the message queue sets a theme Topic, and the message queue sends intelligent voice conversation data to the set theme Topic; the data center subscribes to the theme Topic to obtain intelligent voice conversation data under the theme Topic; a real-time statistics engine in the data center counts the hit times of each branch of the robot speech and the hit times of the knowledge base in the intelligent speech dialogue data in real time, and synchronously updates the statistics data into the database; and the statistical data in the database is fed back to the intelligent voice dialogue system through a preset API after being subjected to data aggregation. The digitalized speech hit rate of the intelligent speech robot facilitates subsequent optimization of speech of the intelligent speech robot.

Description

technical field [0001] The invention relates to the field of intelligent voice dialogue, in particular to a method for analyzing intelligent voice dialogue data operation. Background technique [0002] With the rapid development of computer and artificial intelligence technology, intelligent voice dialogue has been widely developed and applied, intelligent robots have increasingly entered people's life and work, and the application field of robots in life and work is becoming more and more Widely, the era of intelligent robots is coming, and a large number of them are widely used in society. [0003] At present, ASR (Real-time Speech Recognition) and NLP (Natural Language Understanding) can be used to realize real-time understanding of human voice and language by machines, conduct AI intelligent communication in customer service, sales and other scenarios, and conduct large-scale corpus training on human voice language. Under this condition, a recognition model with better ...

AI Technical Summary

Problems solved by technology

[0004] The basis of the current intelligent voice dialogue is composed of speech skills, knowledge base, etc. A set of speech skills may be used thousands of times a day, but there are no quantifiable indicators for the effect of this set of speech skills and whether the design is reasonable, which is not conducive to The continuous improvement of speech skills is not conducive to the improvement of the efficiency of intelligent language robot dialogue

[0005] Most of the existing technical solutions for quantifying the speech skills of intelligent robots focus on the statistics of overall data. For example, the quantitative indicators applied to outbound telephone robots mainly include: connection rate, call duration, etc., and lack of information on the operation process of intelligent voice dialogue. Perform effective quantitative statistical analysis on the words involved

Images