System and method for carrying out intelligent retrieval on writing materials by utilizing semantic fuzzy search

Abstract

The invention discloses a system and a method for carrying out intelligent retrieval on writing materials by utilizing semantic fuzzy search. According to the system, targeted high-authority and high-quality material collection can be performed according to factors such as test scenes, purposes and formats, so that the quality required by a user can be better guaranteed; by introducing a semanticsearch module, the semantic problem is fully considered, the semantic similarity between statements is efficiently judged in a layered semantic similarity calculation mode, and the search matching accuracy is greatly improved; according to the preprocessing method of the sentence vector of the material library, the content of each material in the material library is subjected to simple sentence splitting according to the terminator, grouping is carried out according to the length of the request character field, and the operation rate is remarkably improved; according to an expected statement input by a user, sorting is carried out according to the semantic association degree, multiple retrieval results are displayed at the same time, the user is allowed to view a material original text according to all output results, and the user can have more comprehensive selection.

Description

technical field [0001] The invention relates to the related fields of combining semantic fuzzy search and intelligent retrieval of writing materials, in particular to a system and method for intelligently retrieving writing materials by using semantic fuzzy search. Background technique [0002] In today's society, network information is increasing day by day, how to quickly and effectively find the information that users really need from a large amount of information has become a hot research topic. To put it simply, network information is mainly composed of a large amount of text, and it is the core of the patent of this invention to accurately retrieve the really useful information from the large amount of text. The technology mainly involved in the present invention is fuzzy search, that is, to complete the text matching task in a large amount of text information. Initially, the text matching mainly uses BF (BruteForce), RK (Robin-Karp), KMP (Knuth-Morris-Pratt), Algorit...

AI Technical Summary

Problems solved by technology

[0004] 1. The search space is too large. With the existing technology, the recommended search for good sentences and model essays is often based on the subject or command as the search condition, and the entire length of the search cannot be fuzzy searched for sentence-level feature capture. Therefore, the title The variety of content and propositions makes it difficult for users to find the content they want, or they need to go through a complicated screening process to get a small part of the content they need from many texts, which will consume a lot of time and energy;

[0005] 2. The accuracy of search matching is low, it cannot support semantic association well, and it cannot solve the problem of semantic deviation of key sentences caused by the context in the text, which will reduce the recall rate of search

Moreover, when analyzing semantic similarity, a relatively single similarity measurement index is used, and the similarity between semantics cannot be accurately calculated, that is, the correlation between semantics cannot be efficiently measured, which will lead to the success of the search. The rate will drop, and many search results are empty, but there may actually be text sentences that meet the user's needs;

[0006] 3. The search and matching speed is slow. When searching for complex or long sentences, more violent methods will be used, such as enumeration and posting, to process the text, resulting in low efficiency, slow matching speed, and long time-consuming

Images