Machine translation model performance detection method and related equipment

Abstract

The embodiment of the invention discloses a machine translation model performance detection method and related equipment, and the method comprises the steps: obtaining a source text of a first language type, and obtaining a target text of a second language type corresponding to the source text; replacing the second target words in the source text with a plurality of first target words of the first language type to obtain a plurality of candidate texts; translating the plurality of candidate texts through a machine translation model to obtain a first translated text of the second language type corresponding to each candidate text; translating the source text through a machine translation model to obtain a second translated text of the second language type; and detecting a translation performance index of the machine translation model according to a first similarity between the first translation text and the target text and a second similarity between the second translation text and the target text. And the performance detection accuracy of the machine translation model is improved.

Description

technical field [0001] The present application relates to the technical field of natural language processing, in particular to a method for detecting the performance of a machine translation model and related equipment. Background technique [0002] Machine translation technology refers to the technology of using computer equipment to translate the original text of one natural language (generally called the source language or source) into another natural language (generally called the target language or target). Since machine translation can be done autonomously by computer equipment, it can handle a large amount of translation work in a relatively short period of time compared with human translation. [0003] In the prior art, machine translation can be realized through a machine translation model, for example, a statistical machine translation model (Statistical Machine Translation, SMT). The machine translation model generally relies on a word or phrase alignment table (P...

AI Technical Summary

Problems solved by technology

[0003] In the prior art, machine translation can be realized by a machine translation model, for example, a statistical machine translation model (Statistical Machine Translation, SMT). The machine translation model generally relies on a word or phrase alignment table (Phrase Table) to translate the source language into The target language, and when judging the performance of the machine translation model, the target language translated by the machine translation model is compared with the standard language corresponding to the source language, and the structure of the machine translation model, the number of parameters, and the volume of training data The increasing complexity and scale make it impossible to fully recognize the source of performance bottlenecks through the internal structure of the machine translation model, and it is also impossible to effectively control the acceptability of the prediction of the machine translation model relative to the user after any input is given. Therefore, the overall The comparison method makes the accuracy of the judgment of the performance of the machine translation model lower

Images