There are two necessary conditions to realize full-text retrieval: first, the final information itself needs to be input into the computer, which seems simple, but in fact it contains great workload; Second, the corresponding software support is needed. Full-text information includes not only the text information of books and periodicals, but also various information materials such as pictures, sound and video information. Even if only the text information of books, periodicals and various materials is input into the computer, the amount of information is very large.
At present, full-text retrieval is mainly realized in the following ways:
① Use freely specified retrieval items (such as keywords, character strings, etc.). ) directly compare with full-text original data at high speed for retrieval;
(2) carrying out position scanning on each retrieval item in the text content, and then sorting to establish a reverse document with the discrete code of each retrieval item as an entry;
③ Using hypertext model to establish full-text database and realize hypertext retrieval.