Development of Semantic Collation System in a Special Field
Zheng He Feng · Journal of Southwest Jiaotong University · 2003
This paper presents the design principle and implementation of a semantic collation system to automatically detect political errors in newspapers and magazines before their publication. The sentences are cut into words, followed by detecting sensitive words and terms. The errors of some words and terms are picked up easily by their meanings; and that of the others are determined by fuzzy matching of the whole sentence where the words or terms locate with the knowledge database.