A search terminal control method and search terminal

Abstract

The present application relates to a search terminal control method and a search terminal. Based on an open-source vector database, a vector index subfile is established according to a vector input by a client to be written and stored in an external memory, and then the vector index subfile is mapped The memory object of ElasticSearch is written into the internal memory, so that the ElasticSearch search engine can meet the needs of scalar search and vector search at the same time. merge. When the overall data size of all vector index sub-files under an index name reaches the preset data size, by merging all vector index sub-files under this index name, the fragmented vector index sub-files can be evicted and eliminated to prevent memory loss OOM makes the ElasticSearch search index stable and controllable.

Description

technical field [0001] The present application relates to the technical field, and in particular, to a method for controlling a search terminal and a search terminal. Background technique [0002] The US Internet Data Center pointed out that the data on the Internet will grow by 50% every year and double every two years, and more than 90% of the data in the world is only generated in the last few years. At present, with the upsurge of big data, users are more and more willing to search for data, and the scale of data that can be searched is also increasing. It is difficult for traditional search engines to meet users' massive data search needs. [0003] Conventional search methods generally include scalar search and vector search. A scalar is a single index, a one-dimensional vector such as gender, age, etc. A vector is multi-index and has multiple dimensions (dimensions greater than or equal to 1), such as polynomial coefficients [w1,w2,...,wn]. Scalar search is based o...

AI Technical Summary

Problems solved by technology

[0005] Based on this, it is necessary to provide a search terminal control method and a search terminal for the problem that there is no search engine that can support scalar search and vector search at the same time, and can meet the needs of massive data search.

Images