Parallel hardware for faster morphological analysis

1998

This paper presents an overview of research on computational Arabic morphology which i s d o n e i n t h e W est, as well as a new multi-tape model which consits of three components: A lexical component which mapps the autosegmental morphemes of Arabic (e.g., pattern, root and vo c a l i s m i n s t e m s ) t o a surface form. A rewrite-rules component t a k es care of conditional changes. A morphotactic component p r o vides for morphotactic parsing.

Arabic morphology poses special challenges to computational natural language processing systems. Its rich morphology and the highly complex word formation process of roots and patterns make computational approaches to Arabic very challenging. In this paper we present an approach for morphological analysis and generation of Modern Standard Arabic (MSA). Our approach is based on Arabic morphological automaton technology. We take the special representation of Arabic morphology (root and scheme) to construct a set of morphological automaton which will be used directly in developing a system for Arabic morphological analysis and generation. Our approach for Arabic morphological analysis and generation can be used in different Arabic NLP applications such as Machine Translation (MT) and Information Retrieval (IR).

Research in Computing Science, 2014

Natural language processing applications are based on the morphology part. So they should meet some criteria in order to satisfy the required functionality. Assessing and evaluating of Arabic morphological systems depend on the input words and resulted output according to a predefined criteria to measure and analyze given system in order to study its weakness and strength, trying to find an Arabic morphological analyzer free from all mistakes. In this paper we developed the precise assessment criteria for Arabic morphological analyzers to be applied on a given Arabic morphological analyzers and stemming algorithms by voting, after running them on the sample documents selected as the gold standard.

Indonesian Journal of Electrical Engineering and Computer Science

This article presents a systemof a morphological analyzer of the Arabic language, by integrating several approaches and the viterbi algorithm. First approach is based on database for all thesurface patterns in the Arabic language, second approach is Buckwalter Arabic morphological analyzer and the last approach is based on finite state automaton. With the integration of correspondence tables between affixes in these approaches. The combination between these approaches in our analyzer is very important. Our analyzer is tested on a morphological corpus of 200,000 words, which generalize the words of the Arabic language. The effectiveness of the proposed approaches is demonstrated experimentally and the results obtained are comparable to the state of the art. Moreover, it shows the interest and the advantages of integrating these approaches are to improve our morphological analyzer.

2002

Arabic morphology represents a special type of morphological systems. It is generally considered to be of the nonconcatenative type which depends on manipulating root letters in a nonconcatenative manner. In addition to prefixation and suffixation, inflectional and derivational processes may cause stems to undergo infixational modification in the presence of different syntactic features as well as certain stem consonants. The basic problem, then, is the large number of variants that must be analyzed or generated. In this paper, we seek to reduce the complexity of Arabic morphology using the lexeme-based morphology theory to represent the linguistic resources and MORPHE as a computational tool to implement them. We show that the space of rules can be kept small if we consider the stem as the phonological domain of realisation rules. The reduction in the number of rules keeps the system small and also increases its understandability and maintainability. We primarily focus on generation of verbs and broken plurals.

Annamalai University, 2021

This thesis investigates noun formation in Arabic from a computational point of view. It is about the computational morphological generation and analysis of Arabic nouns. The study first gives a descriptive analysis of Arabic noun morphology based on the stem-based approach, which satisfies the linguistic description and the computational formalization. Both derivational and inflectional systems are discussed in detail. Morphotactics, morphophonemics and the orthography of Arabic nouns are also addressed. The study then presents a computational implementation of Arabic nouns based on the rule-based approach to computational morphology. The overall system is implemented using the NooJ toolkit that supports both finite-state automata (FSA) and pushdown automata (PDA). The system of morphological generation and analysis consists of three components: a lexicon, morphotactics, and rules. The lexicon component lists lexical items (indivisible words and affixes), the morphotactics component encodes constraints on morphemes ordering, and the rules component maps the lexical representations to the surface representation and vice versa. Morphological, morphophonemic, orthographic, and other rules are encoded as two-level rules. The input of the system consists of a main editable lexicon of lemmas taken from three sources: Buckwalter Arabic morphological analyzer lexicon, and Arramooz and Alghani Azzahir machine-readable dictionaries. The output of the system is a full annotated lexicon of inflected noun forms (compiled into a single type of finite-state transducers (FSTs)). This generated lexicon is subsequently used for morphological analysis. Finally, the study presents the evaluation of the system. Three common measures are used to evaluate the performance of the system: accuracy, precision, and recall. The evaluation task consists of conducting two empirical experiments. The first experiment evaluates the system performing morphological analysis on diacritized Arabic words. The system’s performance using diacritized Arabic words for accuracy, precision, and recall, is 90.4%, 98.3%, and 88.9%, respectively. The second experiment evaluates the system on undiacritized words. The obtained results of this experiment in terms of accuracy, precision, and recall are 94.7%, 96.7%, and 91.6%, respectively. The average of the measures for the two experiments has been also calculated. The average of the performance values in terms of accuracy, precision, and recall is respectively 92.55%, 97.5%, and 90.25%. In general, the results are promising and they show the ability of the system to deal with different types of Arabic texts, diacritized and undiacritized texts. This system can provide a detailed analysis and morphological tags for nouns in Arabic text corpora. It divides the analyzed word into three parts (i.e., proclitics/prefixes, stem, and suffixes/enclitics) and gives each part a detailed morphological feature tag or possibly multiple tags if the parts have multiple clitics or affixes.

Arxiv preprint arXiv:1101.5494, 2011

Arabic morphological analysis is one of the essential stages in Arabic Natural Language Processing. In this paper we present an approach for Arabic morphological analysis. This approach is based on Arabic morphological automaton (AMAUT). The proposed technique uses a morphological database realized using XMODEL language. Arabic morphology represents a special type of morphological systems because it is based on the concept of scheme to represent Arabic words. We use this concept to develop the Arabic morphological automata. The proposed approach has development standardization aspect. It can be exploited by NLP applications such as syntactic and semantic analysis, information retrieval, machine translation and orthographical correction. The proposed approach is compared with Xerox Arabic Analyzer and Smrz Arabic Analyzer.