No abstract

No abstract

In this paper, we present a Dialect Identification system (ArbDialectID) that competed at Task 1 of the MADAR shared task, MADAR Travel Domain Dialect Identification. We build a coarse and a fine grained identification model to predict the label (corresponding to a dialect of Arabic) of a given text. We build two language models by extracting features at two levels (words and characters). We firstly build a coarse identification model to classify each sentence into one out of six dialects, then use this label as a feature for the fine grained model that classifies the sentence among 26 dialects from different Arab cities, after that we apply ensemble voting classifier on both subsystems. Our system ranked 1st that achieving an f-score of 67.32%. Both the models and our feature engineering tools are made available to the research community.

Recently, the need has increased for an effective and powerful tool to automatically summarize text. For English and European languages an intensive works have been done with high performance and nowadays they look forward to multi-document and multi-language summarization. However, Arabic language still suffers from the little attentions and research done in this filed. In this paper, we propose a model to automatically summarize Arabic text using text extraction. Various steps are involved in the approach: preprocessing text, extract set of features, classify sentence based on scoring method, ranking sentences and finally generate an extracted summary. The main difference between the proposed system and other Arabic summarization systems are the consideration of semantics, entity objects such as names and places, and similarity factors in our proposed system. The proposed system has been applied on news domain using a dataset osbtained from Local newspaper. Manual evaluation techniques are used to evaluate and test the system. The results obtained by the proposed method achieve 86.5% similarity between the system and human summarization. A comparative study between our proposed system and Sakhr Arabic online summarization system has been conducted. The results show that our proposed system outperforms Shakr system.