Search question-answering system based on pre-training

The invention discloses a search question-answering system based on pre-training. The search question-answering system comprises a noise judgment module, a QA question-answering module, a knowledge matching module and a response output module, the noise judgment module judges whether the user problem belongs to noise or not; the QA question and answer module comprises a rule input unit and a ruleanalysis unit. The knowledge matching module is used for indexing questions and knowledge in a question and answer library and performing similarity sorting, knowledge indexes comprise an inverted index mode and an Annoy index mode, the response output module is used for outputting responses, and the responses comprise four types of similar question lists, accurate answers, no answers and recommended popular questions. The problems of knowledge generalization migration, noise judgment and QA customization can be effectively solved, and the question and answer efficiency is greatly improved while the user experience is