Novel textual entailment technique for the Arabic language using genetic algorithm. (July 2021)

• Record Type: Journal Article
• Authors: Alhijawi, Bushra; Awajan, Arafat
• Is Part Of: Computer speech & language. Volume 68(2021)
• Publication Date: 2021-07
• Subjects: Recognizing textual entailment -- Text similarity -- ArbTEDS -- Genetic algorithms -- Arabic natural language processing; Speech processing systems -- Periodicals; Automatic speech recognition -- Periodicals; Computers -- Periodicals; Linguistics -- Periodicals; Speech-Language Pathology -- Periodicals; Traitement automatique de la parole -- Périodiques; Reconnaissance automatique de la parole -- Périodiques; Automatic speech recognition; Speech processing systems; Electronic journals; Periodicals; 006.454
• DOI: 10.1016/j.csl.2021.101194
• Languages: English
• ISSNs: 0885-2308

Abstract

Highlights: We propose a novel textual entailment recognition model that employs a genetic algorithm. The textual entailment is considered as an optimization problem. We use GA to find the optimal weights and threshold. We apply the new model in Arabic text pairs. We compare the new model with other methods on ArbTED dataset. The comparative results demonstrate the superiority of the proposed model. Abstract: This paper presents a textual entailment (TE) model that considers entailment as an optimization problem. The proposed TE model employs a genetic algorithm to derive an optimal similarity function and correlated entailment judgment threshold. The similarity function is formulated through a linear combination of text similarity measures and weights. Two text similarity measures are considered: cosine and the longest common substring. These text similarity measures are computed for each text pair. The weights represent the importance of the considered text similarity measures for generating an entailment judgment. The weights and correlated judgment thresholds are obtained by the genetic algorithm. Several experiments are conducted using the ArbTED dataset to evaluate the performance of the proposed model. Comparative results demonstrate the superiority of the proposed model. On average, the model achieved a 16% improvement in terms of accuracy. Furthermore, the average recall and precision values were 72.7% and 72.3%, respectively.