摘要 |
<p>A method, device and system for natural language processing on the basis of semantic recognition, comprising: using punctuation to parse words in a text into character strings, and extracting language linear structures and language chunks from said strings (101); listing in reverse order the extracted language linear structures and language chunks (102); creating a language linear structure subindex and a language chunk subindex, and then combining both subindices into a single complete index (103); extracting language linear structures and language chunks from search strings input by a user, and feeding back to the user information matching same (104). The present invention uses a technical means of analyzing language structures + key words to grasp precisely the true meaning of the information therein, and accurately provides the user with the information needed.</p> |