원문정보
초록
영어
We present an automatic summarization of Arabic texts. In this context, two methods for automatic summarization of Arabic documents are proposed and comparatively analyzed. The first method is based on rhetorical structure theory RST that uses linguistic knowledge. The second method is based on digital learning theory that relies on mathematical rules. Both methods are implemented using the Arabic summarization RST ASRST and Arabic summarization digital learning ASDL systems. Thereafter, results obtained by the two systems are presented and a comparative study between the obtained results is conducted to highlight the advantages and limitations of each method. Evaluation results showed that the numerical approach has better performance than the symbolic approach, particularly when the texts are short. By contrast, the symbolic approach provides better result than the numerical approach when the texts are long.
목차
1. Introduction
2. Summary Based on RST
2.1. Rhetorical Relations Used in the Analysis
2.2. Proposed Method
2.3. Arabic Rhetorical Relation
3. Proposed Numerical Method
3.1. Presentation
3.2. Main Stages of the Proposed Method
4. Evaluation
5. Discussion of Results
References