English writing aided semantic matching method and device

Abstract

The invention relates to an English writing aided semantic matching method and device. The matching method comprises the following steps: S1, constructing an example sentence corpus; s2, constructing a semantic matching model training data set; s3, constructing a semantic matching model of the multi-view attention mechanism; and S4, training a semantic matching model. According to the English writing aided semantic matching method and device provided by the invention, a remarkable effect is achieved on several public semantic matching data sets, richer interaction matching information can be extracted, the relation and the difference between sentence pairs can be better captured, global matching information can be better obtained, and the invention has a great application prospect in related fields of text matching.

Description

technical field [0001] The invention relates to the technical field of natural language processing, in particular to a semantic matching method and device for English writing assistance. Background technique [0002] As the most widely spoken language in the world, English plays an important role in daily academic communication and business dealings. As an important channel of formal communication, writing has become second only to spoken language and a barrier between many non-foreign language workers and the foreign world. The writing level of a writer can highlight his proficiency in the language. To write high-quality articles, he must have a lot of reading experience and knowledge reserves. However, for most non-English-speaking students, due to the lack of English reading experience and knowledge reserve, writing high-quality English articles is by no means an easy task. In the process of writing, they often organize the content of the article in their native languag...

AI Technical Summary

Problems solved by technology

However, through the analysis and research of the existing auxiliary writing systems, it is not difficult to find that most of them still use very traditional methods, and the effect is relatively limited

The traditional statistical machine learning stage requires a lot of manual experience and time for early data labeling and mid-term feature extraction, and the experimental results for complex tasks are not ideal, and it is not good for text representation and understanding of semantic information

Moreover, problems such as the curse of dimensionality and data sparsity caused by traditional text representation methods have affected the development of the field of natural language processing.

Images