Nowadays, the amount of Arabic documents has increased significantly in different domains, such as news articles, emails, business summary, biomedicine, web sites and social media documents. Some databases have increased in its size to terabyte. Multi-document summarization is the method of creating a summary of a group of interrelated documents. Therefore, the rise of the desire for Arabic multi documents text summarization (at the instant rates possible, coherent, grammatical and meaningful

more »

... ntences) is increased. Recently, many efforts on multi-document text summarization that is related to the English language have been performed. Arabic multi-document summarization is remained on its early stages. Consequently, the researchers in this paper propose an Arabic Multi-Document Text Summarization (AMD-TS) model based on parallel computing techniques. This model of Arabic text summarization could effectively and rapidly summarize Arabic multi-documents in real time. A conceptual framework is proposed based on published researches dealing with text summarization techniques of different languages. The proposed model creates an accurate, coherent and complete Arabic multi-document text summarization model. The dataset that is used in the investigation stage is derived from different domains, such as education, sports and politics. This dataset contains texts of various sizes. The experiments are then designed to be on specific domain (news articles domain). In order to increase the summarization process efficiency and performance, the researchers in this paper use parallel computing. The model covers the deficiency of Arabic Automatic Summarization Systems (ASS) by enhancing the final summary.