Design of Corpus Distributed Retrieval System Based on Elasticsearch Search Engine

Information retrieval is the best practice to solve information data overload. Elasticsearch is a distributed, high-performance elastic full-text search engine based on Lucene extension. Based on Elasticsearch framework, this paper studies the distributed retrieval of corpora to meet the needs of the research and use of corpora in the information age. The research content mainly includes four aspects: First, the Elasticsearch architecture is studied, including the functions and data operations of each layer. Then, the N-Gram linguistic statistical model is studied, based on the basic N-Gram model, Markov hypothesis is introduced to improve it, and the parameter estimation and data smoothing of the N-Gram model are studied. Then, a corpus distributed retrieval module is designed, including index creation, retrieval process design and highlighting design. Preferably, the system is tested and validated, including building test environments, functional testing, and compatibility testing. The test results show that the system achieves the expected goal and has high retrieval efficiency and accuracy. Research has found that a corpus distributed retrieval system designed based on Elasticsearch search engine has fast retrieval speed, good practicality, and reliability.