Mohammed Attia

4 paper cards in Mova's library · 2011–2017

Mohammed Attia wrote 4 of the papers Mova reads, in 2011–2017, most of them on Treebanks, Universal Dependencies; often with Ahmed Abdelali, Hamdy Mubarak, Kareem Darwish.

The profile is the library's match of the name; two people with one name are not always told apart.

Papers

CoNLL 2017 Shared Task: Multilingual Parsing from Raw Text to Universal Dependencies (2017)
The shared task asks systems to parse raw text, with no gold tokenization, sentence splitting or tags, into UD trees for 49 languages.
A Neural Architecture for Dialectal Arabic Segmentation (2017)
Egyptian Arabic in tweets has no standard spelling and few annotated resources, and the clitics attached to its verbs are missed by MSA tools.
Arabic Multi-Dialect Segmentation: bi-LSTM-CRF vs. SVM (2017)
The paper segments four Arabic dialects (Egyptian, Levantine, Gulf, Maghrebi) from raw Twitter text. Only about 350 manually segmented tweets per dialect are used for training, with 5-fold cross-validation and a 70/10/20 split. Two approaches are compared. An SVM ranker generates all segmentations of a word from a…
An Open-Source Finite State Morphological Transducer for Modern Standard Arabic (2011)
Attia et al. build AraComLex, an open-source finite-state morphological analyzer for Modern Standard Arabic, compiled with Foma from a lemma-based lexc lexicon and about 130 alteration rules. They grow a seed lexicon of 10,799 lemmas to 30,587 lemmas. The additions come from a billion-word corpus pre-annotated with…