Original Articles
Shitao Yan and Wenqiang Tian
Abstract
In order to improve the effect of query accuracy in intelligent question-answering system, the sentence similarity calculation model has been a research focus in this field. Based on the analysis of overall implementation framework in intelligent question-answering system, this paper proposes a new type of sentence similarity calculation model based on a linear regression algorithm, which combines the advantages to do sentence similarity computing based on keywords characteristics and semantic characteristics, according to the characteristics to calculate the sentence similarity and obtain a high accuracy