An efficient single document Arabic text summarization using a combination of statistical and semantic features

Extractive Multi-Document Arabic Text Summarization Using Evolutionary Multi-Objective Optimization With K-Medoid Clustering

“…According to Nenkova, A. et al [23], machine learning approaches are well suited for single document more than multi-document summarization. Moreover, studies have shown the effectiveness of this approach [24]. However, this approach needs labeled data (training dataset), and the creation of such dataset is time-consuming task.…”

“…In addition, it includes statistical and semantic features. It is worth mentioning that the selection is based on our observations as well as the studies, experiments, and results made in other related works [3], [4], [24], [88]. The features used in the proposed approach are explained below.…”

“…For example, the first sentence in a document is the most important sentence [91]. We model the sentence location score based on [24] as:…”

Enhancement of Single Document Text Summarization using Reinforcement Learning with Non-Deterministic Rewards

“…The study is carried out in this regard. In paper [1], an automatic extractive Arabic single document summarizing method for producing a informative summary from documents. It also uses two summarizing techniques score-based and supervised machine learning for generating summaries.…”

“…The Table 2 & Table 3 shows the ROUGE evaluation measure attained from various systems such as baseline, reinforcement learning and reinforcement learning with QA corpus (proposed) for benchmark dataset. The following Table 4 shows the performance analysis of machine-generated summary with the golden standard summary generated by human [1]. Standard Classifiers such as Support Vector Machines (SVM), Naï ve Bayes, Neural Networks are considered for evaluation.…”