PROPAGATING TEMPORAL RELATIONS OF INTERVALS BY MATRIX
Authors:
Shichao Zhang; Chengqi Zhang
DOI:
10.1080/088395102753365771
Publication Frequency:
10 issues per year
Subjects:
Artificial Intelligence;
Computer Science (General);
Information & Communication Technology (ICT);
Number of References: 45
Formats available:
PDF
(English)
View Article:
View Article (PDF)
Abstract
Traditional temporal relations propagating is based on Allen's Interval Algebra. This paper proposes an alternative method to propagate temporal relations among intervals, in which 5
5 matrices are used to represent temporal relations of intervals. Hence, the propagation of temporal relations is transformed into a numerical computation. For efficiency, we use the special values of the thirteen matrices to determine the possible temporal relations between two given intervals by using only the final resultant matrix so as to optimize the propagation. To evaluate the utility of the proposed technique, we have implemented the matrix representation in Java. The experimental results demonstrate that the approach is efficient and promising.
|
| view references (45) : view citations |

Download Citation
5 matrices are used to represent temporal relations of intervals. Hence, the propagation of temporal relations is transformed into a numerical computation. For efficiency, we use the special values of the thirteen matrices to determine the possible temporal relations between two given intervals by using only the final resultant matrix so as to optimize the propagation. To evaluate the utility of the proposed technique, we have implemented the matrix representation in Java. The experimental results demonstrate that the approach is efficient and promising.
CiteULike
Del.icio.us
BibSonomy
Connotea