【摘要】：針對現有度量方法中考慮因素不夠全面和因子權重計算依據經驗確定的不足,提出粒子群優(yōu)化BP神經網絡(PSO-BP)的地理本體概念語義相似度度量模型。該模型利用本體屬性、本體結構和語義關系的相似度,結合權重信息計算概念的綜合相似度;同時,利用粒子群算法優(yōu)化的BP神經網絡獲取因子權重,避免現有方法中因子權重確定的人為主觀干擾。最后,從基礎地理信息概念中提取出200組樣本,用其中190組作為訓練集,對神經網絡模型進行訓練,以獲取權重;剩余10組作為測試集。將該模型和幾種常用算法進行對比,通過分析測試集的各算法求解結果和專家判定結果之間的相關系數,結果表明該模型計算地理本體概念的相似度更為準確,符合人類認知特性,效果更好。
[Abstract]:In view of the fact that the factors in the existing measurement methods are not comprehensive enough and the calculation of factor weights is determined by experience, a semantic similarity measurement model of geographical ontology concept based on particle swarm optimization BP neural network (PSO-BP) is proposed. The model uses the similarity of ontology attribute, ontology structure and semantic relationship, combined with the weight information to calculate the comprehensive similarity of the concept. At the same time, the BP neural network optimized by particle swarm optimization algorithm is used to obtain the factor weight to avoid the artificial subjective interference determined by the factor weight in the existing methods. Finally, 200 groups of samples are extracted from the concept of basic geographic information, 190 of which are used as training sets to train the neural network model to obtain the weight, and the remaining 10 groups are used as the test set. The model is compared with several common algorithms, and the correlation coefficients between the results of each algorithm and the results of expert decision are analyzed. The results show that the model is more accurate in calculating the similarity of geographical ontology concepts, in line with the cognitive characteristics of human beings, and the effect is better.
【作者單位】： 信息工程大學;海軍出版社;
【分類號】：TP391.1;TP18

【參考文獻】

相關期刊論文 前10條

1 馬雷雷;梁汝鵬;李宏偉;連世偉;周海;;一種基于描述邏輯的空間語義相似性計算方法[J];測繪科學技術學報;2015年02期

2 楊娜娜;張青年;牛繼強;;基于本體結構的空間實體語義相似度計算模型[J];測繪科學;2015年03期

3 劉一明;胡卓瑋;趙文吉;王志恒;;基于BP神經網絡的區(qū)域貧困空間特征研究——以武陵山連片特困區(qū)為例[J];地球信息科學學報;2015年01期

4 潘潤秋;馬小淞;劉s，

本文編號：2501514