MADAMIRA: A Fast, Comprehensive Tool for Morphological Analysis and Disambiguation of Arabic
By Arfath Pasha, Mohamed Al-Badrashiny, Mona T. Diab, Ahmed El Kholy, R. Eskander, Nizar Habash, Manoj Pooleery, Owen Rambow, Ryan Roth. 2014
692 citations recorded by the source corpus.
In this paper, we present MADAMIRA, a system for morphological analysis and disambiguation of Arabic that combines some of the best aspects of two previously commonly used systems for Arabic processing, MADA (Habash and Rambow, 2005; Habash et al., 2009; Habash et al., 2013) and AMIRA (Diab et al., 2007). MADAMIRA improves upon the two systems with a more streamlined Java implementation that is more robust, portable, extensible, and is faster than its ancestors by more than an order of magnitude. We also discuss an online demo (see http://nlp.ldeo.columbia.edu/madamira/) that highlights these aspects.
Subfields: Artificial Intelligence, Computational Theory and Mathematics, Computational Linguistics & Semantics
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Artificial Intelligence
- Yonatan Belinkov
- Nadir Durrani
- Hassan Sajjad
- Fahim Dalvi
- Philipp Koehn
- James R. Glass
- Percy Liang
- Graham Neubig
In Computational Theory and Mathematics
- Yonatan Belinkov
- Nadir Durrani
- Nizar Habash
- Adam Lopez
- Alexander M. Fraser
- Marco Turchi
- Matteo Negri
- Marcin Junczys-Dowmunt
In Computational Linguistics & Semantics
- Yonatan Belinkov
- Nizar Habash
- Adam Lopez
- Hassan Sajjad
- Nadir Durrani
- Owen Rambow
- Clara Vania
- Fahim Dalvi
Citations in the corpus
- Automatic Gender Identification and Reinflection in Arabic (2019, methodology): “For all the collected and created corpora (manually translated, synthetic and machine translated), we generated a parallel morphologically analyzed feature corpus using the MADAMIRA Arabic analysis toolkit (Pasha et al., 2014).”
- An Empirical Analysis of NMT-Derived Interlingual Embeddings and Their Use in Parallel Sentence Identification (2017, methodology): “A second system called S1-l is trained on lemmatised sentences using the MADAMIRA lemmatiser for Arabic, and the IXA pipeline (Agerri et al., 2014) for 2http://commoncrawl.org 3http://www.casmacat.eu/corpus/ news-commentary.html 4https://sites.google.com/site/ iwsltevaluation2016/mt-track English an”
- Challenging Language-Dependent Segmentation for Arabic: An Application to Machine Translation and Part-of-Speech Tagging (2017, methodology): “Here we use the state-of-the-art Arabic segmenter MADAMIRA and Farasa as our baselines.”
- The Interplay of Variant, Size, and Task Type in Arabic Pre-trained Language Models (2021, methodology): “Dataset We fine-tuned our models on three different POS tagging datasets: (1) the Penn Arabic Treebank (PATB) (Maamouri et al., 2004) which is in MSA and includes 32 POS tags; (2) the Egyptian Arabic Treebank (ARZATB) (Maamouri et al., 2012) which is in Egyptian (EGY) and includes 33 POS tags; and (3”
- From Characters to Words to in Between: Do We Capture Morphology? (2017, methodology): “4We only experimented with Arabic since MADAMIRA disambiguates words in contexts; most other analyzers we found did not do this, and would require additional work to add disambiguation.”
- FCICU at SemEval-2017 Task 1: Sense-Based Language Independent Semantic Textual Similarity Approach (2017, methodology): “Arabic tokens are lemmatized using Madamira 6 (Pasha et al., 2014).”
- On internal language representations in deep learning: an analysis of machine translation and speech recognition (2018, methodology): “corpora: MADAMIRA [272] for Arabic POS and morphological tags, Tree-Tagger [296] for Czech and French POS tags, LoPar [297] for German POS and morphological tags, and MXPOST [283] for English POS tags.”
- Building an Arabic Machine Translation Post-Edited Corpus: Guidelines and Annotation (2016, methodology): “MADAMIRA uses a morphological analyzer to produce, for each input word, a list of analyses specifying every possible morphological interpretation of that word, covering all morphological features of the word.”
- Farasa: A Fast and Furious Segmenter for Arabic (2016, methodology): “Both text and analyses are passed to a feature modeling component, which applies SVM and language models to derive predictions for the word segmentation (Pasha et al., 2014).”
- The MADAR Arabic Dialect Corpus and Lexicon (2018, methodology): “For Arabic, we use MADAMIRA (Pasha et al., 2014) to tokenize words into the D3 scheme, which separates all clitics from the basewords.”
- A Parallel Corpus for Evaluating Machine Translation between Arabic and European Languages (2017, methodology): “We tokenized the Arabic content using the MADAMIRA toolkit (Pasha et al., 2014) with”