Enterprise internal search engine method based on large language model

• Main Authors: MA GUICHENG, ZHENG WENQIAN, GUO FENG, HUANG XIAOFENG, WU ZECHUAN, ZHANG SHIQING
• Format: Patent
• Language: Chinese; English
• Published: 19.09.2023
• Subjects: CALCULATING; COMPUTING; COUNTING; ELECTRIC DIGITAL DATA PROCESSING; PHYSICS

The invention provides an enterprise internal search engine method based on a large language model, and relates to the technical field of enterprise internal search engines. According to the enterprise internal search engine method based on the large language model, firstly, a knowledge base is loaded, and all documents in the knowledge base are loaded; segmenting the knowledge base into independent text states one by one; secondly, vectorizing the knowledge base, mapping the text into vector data, and storing the vector data in vector numerical control; vectorizing user search; performing vectorization matching on a target text according to the search statement; generating a result according to the prompt word; and finally, streaming response is performed to return a result. According to the enterprise internal search engine based on the large language model, document content can be intelligently read and understood, the user context is combined, the performance, efficiency and accuracy of user enterprise in