On the associative memory utilization in English- Arabic natural language processing

Arabic natural language processing (ANLP) consists of developing techniques and tools that can utilize and analyze the Arabic language in both written and spoken contexts. ANLP makes an important contribution to many existing developed systems. It provides Arabic and non-Arabic speakers with helpful and convenient tools that can be used in different domains. Modern ANLP tools are developed using machine learning (ML) techniques. ML algorithms are widely used in NLP because of their high accuracy rate regardless of the robustness of the data that is used and because of the ease with which they can be implemented. On the other hand, the methodology of ANLP applications based on ML involves several distinct phases. It is, therefore, crucial to recognize and understand these phases in detail as well as the most widely used ML algorithms. This survey discusses this concept in detail, shows the involvement of ML techniques in developing such tools, and identifies well-known techniques used in ANLP. Moreover, this survey discusses the characteristics and complexity of the Arabic language in addition to the importance and needs of ANLP.

Arabic Natural Language Processing (ANLP) is an area of Artificial Intelligence (AI) that enables computers to process and understand Arabic text and speech. It powers applications such as translation, sentiment analysis, and speech recognition, opening new horizons. However, despite the worldwide popularity of Arabic, ANLP is lagging. The Arabic language possesses unique linguistic features that complicate computational processing. In this context, our article seeks to explore these challenges and examine Arabic’s inherent characteristics in orthography, morphology, grammar, syntax, and linguistic diversity, which contribute to the complexities of its Natural Language Processing (NLP). Indeed, the main challenges are the diversity of dialects, the morphological and syntactic richness of Arabic, diglossia, and the absence of short vowels. In addition, the scarcity of Arabic resources further complicates NLP efforts. This review can serve as a guide for practitioners in the field of Arabic NLP, whether they are computer scientists or linguists. It also calls on the Arab community scientists to take steps to meet the potential challenges and increase efforts in the field to promote Arabic NLP

The last decade had known a great interest in Arabic Natural Language Processing (NLP) applications. This interest is due to the prominent importance of this 6th most wide-spread language in the wo...

Arabic language is one of the most spoken languages in the world, it’s the official language of many countries and the fourth most used language on the internet. Arabic texts are often written without diacritic marks. However, those marks are important to clarify the sense and meaning of words. Automatic diacritization is the process of assigning diacritics to letters, and it’s an important field in Arabic Natural Language Processing (ANLP). In this work, we try to find the effect of increasing the training dataset on the diacritization error rate (DER) by building a new dataset and concatenating it with the Tashkeela dataset. We trained a deep learning model based on bidirectional long short-term memory BLSTM that transcribes undiacritized sequences of Arabic letters and produces an output sequence of the same length fully diacritized. Our model shows significant results on the new dataset in terms of DER and validation loss.

The legal sector remains distinctive due to the complex language structure and specialized terminology of legal data. This complexity offers considerable contextual information, which demands natural language processing (NLP). The availability of high-quality and well-structured legal datasets is essential for advancing NLP research and applications within the legal field. However, a gap exists within the Arabic legal NLP owing to insufficient research and datasets. To address this gap, we aim to propose an Arabic legal case dataset containing cases, case summaries, relevant keywords, and case categories. The legal case data were obtained from the Board of Grievances website in Saudi Arabia and include 3170 cases distributed across 47 classes. The number of words in these cases varies significantly, ranging from about 100 to nearly 30,000 words per case. Moreover, the number of pages varies, ranging from one page to 80 pages per case. Therefore, this dataset supports various NLP applications, including text categorization, data extraction, sentiment analysis, and summarization, thereby improving task efficiency and decision accuracy in the legal profession.

In the era of Big Data and Data Science, Text analysis within, Natural Language Processing (NLP), suffers from the curse of high dimensionality. The use of NLP in applications such as speech processing, semantic webs, and word processing has become a main element in today’s Artificial Intelligence and Big Data Applications. A natural language parsing system must incorporate three components of natural language, namely, lexicon, morphology, and syntax. As Arabic is highly derivational, each component requires extensive exploitation of the associated linguistic characteristics. Parsing Arabic sentences still has open challenges due to several reasons including the relatively free word order of Arabic, the length of sentences, and the omission of diacritics (vowels) in written Arabic and the frequency of pro-drop phenomena. This research exploits Visual Prolog to provide a scalable platform for Arabic parser and explains the details of the used lexicon and parser and shows the scalability of the system to address more functions.

Metaphor is a literary device that allows us to express a concept in terms of another. In other words, it is based on similarity between concepts. Metaphorical expressions represent a great variety and they are used in conventional metaphors, which we reproduce and comprehend every day, poetic, novel, and Holy Qur’an. The use of metaphor is ubiquitous in natural language text and it is a serious bottleneck in automatic text understanding, and developing methods to identify and deal with metaphors is an open problem in Arabic natural language processing, especially Machine Translation. Due to the complexities involved in metaphor, it semantically influenced the meaning of machine-translated text. This makes metaphor an important research area for computational and cognitive linguistics, and its automatic identification and interpretation is indispensable for any semantics-oriented Arabic natural language processing. In this paper, we present the challenges of Arabic NLP of metaphors, which is very important in developing a computational NLP-based system for in Classical Arabic, Modern Standard Arabic and Dialect Arabic. We also highlight main problems that arises when translating an Arabic metaphor to another language.

Morphologically complex languages such as Arabic pose several challenges in Natural Language Processing (NLP) due to their complexity and token sparsity. Most techniques approach the problem by transforming the words of the language from their sparse surface form representation to a less sparse form before processing. The transformation usually takes the form of a morphological analysis or a morphological segmentation. This dissertation addresses two tasks in Arabic NLP: Statistical Machine Translation(SMT) and Sentiment Analysis. To improve English-Arabic SMT, we apply segmentation on Arabic to decrease token sparsity and enhance the correspondence between tokens of the English and Arabic language. However, due to this segmentation, the translation system is limited to extracting features based on morphemes (partial words) and only outputting morphemes during decoding. Such a system lacks knowledge of the original form of the words. We further improve translation from English to Arabic by combining both segmented and desegmented views of the target language. The system can benefit from segmentation's sparsity reduction and verifies its generation of correct words. We present a language-independent technique to desegmentation that approaches the problem as a string transduction task. We propose a new algorithm that desegments the decoder's search space encoded as a lattice, thus allowing the system to use features from the desegmented view of the search space. We extend the phrase-based statistical machine translation system to allow desegmentation during the decoding process on the fly. In addition, we conduct an experimental study to verify what matters most in morphologically segmented SMT models. Our second task is sentiment analysis, where we resort to Arabic lemmatization to improve sentiment analysis of Arabic tweets and blog posts. We explore translation in the opposite direction, from Arabic into English in order to evaluate the loss of sentiment predictability when Arabic social media posts are translated to English, manually or using an SMT system. We use state-of-the-art Arabic and English sentiment Analysis systems and develop automatically generated Arabic lexicons from lemmatized tweets to improve this task.

Arabic Named Entity Recognition (ANER) systems aim to identify and classify Arabic Named entities (NEs) within Arabic text. Other important tasks in Arabic Natural Language Processing (NLP) depends on ANER such as machine translation, question-answering, information extraction, etc. In general, ANER systems can be classified into three main approaches, namely, rule-based, machine-learning or hybrid systems. In this paper, we focus on research progress in machine-learning (ML) ANER and compare between linguistic resource, entity type, domain, method and performance. We also highlight the challenges when processing Arabic NEs through ML systems.

introduction Share on Introduction to the Special Issue on Arabic Natural Language Processing Authors: K. Shaalan The British University in Dubai The British University in DubaiView Profile , A. Farghaly Monterey Institute of International Studies Monterey Institute of International StudiesView Profile Authors Info & Claims ACM Transactions on Asian Language Information ProcessingVolume 8Issue 4December 2009 Article No.: 13pp 1–3https://doi.org/10.1145/1644879.1644880Published:01 December 2009Publication History 1citation365DownloadsMetricsTotal Citations1Total Downloads365Last 12 Months9Last 6 weeks4 Get Citation AlertsNew Citation Alert added!This alert has been successfully added and will be sent to:You will be notified whenever a record that you have chosen has been cited.To manage your alert preferences, click on the button below.Manage my AlertsNew Citation Alert!Please log in to your account Save to BinderSave to BinderCreate a New BinderNameCancelCreateExport CitationPublisher SiteGet Access

AraVec: A set of Arabic Word Embedding Models for use in Arabic NLP

Computational linguistic and natural language processing automation tasks require text annotated with tags that represent the desired output of the task. The annotation tags serve for training, validation, and evaluation. Arabic morphological analysis, and tags associated with it such as part of speech and gloss tags, is key to Arabic computational linguistics and natural language processing. Several manual and automated tagging tools exist for text. Very few exist that are based on Arabic morphological analysis. In this paper, we present an open source tagging tool with visual interface that enables the construction of annotated Arabic text corpora with automatic morphology-based tags. The tool allows the specification of tags with Boolean formulae where the atomic predicates are match and contain relations between the morphological solution of part of the text and the value of a morphological feature. The tool allows the user to directly enter manual tags, to edit existing tags through a tag sensitive coloring interface, to compare tag sets, and compute accuracy results.

Conventional requirements engineering (RE) is often a manual process that is tedious, time-consuming, susceptible to errors, and hard to scale in case of continuous updates. Recent advances in natural language processing (NLP) transformed the RE process via automation. However, automation works better with structured data and formal requirement formats, limiting the ability to elicit and extract requirements from unstructured data; the rapid growth of digital platforms accentuates the need for treating user posts, comments, opinions and feedback as primary data sources for requirements mining and modeling. However, research conducted on unstructured Egyptian Arabic text is largely underexplored, due to its inherent linguistic challenges. This paper fills a critical gap in Arabic NLP, by providing an automated feature extraction and UML modeling pipeline utilizing state of the art embedding models, density-based clustering algorithm, and Large Language Models (LLMs), specifically engineered for handling Egyptian-Arabic text. The proposed pipeline demonstrates strong positive results, producing quality features and meaningful use case diagrams, and further confirms the feasibility of applying this advanced NLP-based pipeline to real world Egyptian dialectal data.

Large language models (LLMs) demonstrate impressive capabilities across a range of natural language processing (NLP) tasks. However, they are highly sensitive to prompt design, which significantly affects their ability to align outputs with user intent. Poorly crafted prompts can result in misleading or irrelevant responses. Nevertheless, selecting the most effective prompt from several candidates remains an open challenge. Despite the growing importance of prompt engineering, there is no comprehensive framework to systematically evaluate prompts across multiple dimensions, such as similarity, performance, efficiency, and consistency, particularly in scenarios where performance can be traded off against computational cost or consistency. In this study, we propose a novel scoring framework to evaluate handcrafted prompts across four essential dimensions: Similarity, performance, efficiency (measured by latency, input tokens, and output tokens), and consistency. Considering Arabic, a relatively low-resource, morphologically rich language, as a case study, we evaluated this framework on six diverse text classification tasks: Dialect identification, sentiment analysis, offensive language detection, stance detection, emotion detection, and sarcasm detection. Our methodology assesses prompts across multiple LLMs (GPT-4o mini, LLaMA, ALLAM, and Claude 3.5 Haiku), providing valuable insights into model-specific and task-specific performance patterns. Results demonstrate that no single prompt universally excels across all dimensions; rather, optimal prompts vary based on specific task requirements and evaluation priorities. The proposed framework enables the identification of the most effective prompts for each application context while revealing important trade-offs between performance metrics. By addressing the unique challenges of Arabic NLP, this research not only advances prompt engineering for underrepresented languages but also provides a systematic and adaptable methodology for prompt evaluation that can enhance LLM performance across a range of linguistic contexts, diverse domains, tasks, and various model architectures.

Compared with English, Arabic is a poorly-resourced language within the field of corpus linguistics. A lack of sufficient data and research has negatively affected Arabic corpus-based researchers and natural language processing practitioners. Although a number of Arabic corpora have been developed in recent years, the overall situation has improved little. The aim of this paper is twofold. First, it reviews 14 Arabic corpora categorized by their designated purpose, target language, mode of text, size, text date, location, text type/medium, text domain, representativeness, and balance. The review also describes the availability of the reviewed corpora, the presence of tokenization, lemmatization and tagging, and whether there are any tools available to search and explore them. Second, it introduces the King Abdulaziz City for Science and Technology (KACST) Arabic corpus, which was designed and created to overcome the limitations of existing Arabic corpora. The KACST Arabic corpus is a large and diverse Arabic corpus with clearly defined design criteria. It is carefully sampled, and its contents are classified based on time, region, medium, domain, and topic, and it can be searched and explored using these classifications. The KACST Arabic corpus comprises more than 700 million words from the pre-Islamic era to the present day (a period covering more than 1,500 years), collected from 10 diverse mediums. Each text has been further classified more specifically into domains and topics. The KACST Arabic corpus is freely available to explore on the Internet ( http://www.kacstac.org.sa ) using a variety of tools.