nmatheg

nmatheg نماذج an easy straregy for training Arabic NLP models on huggingface datasets. Just specifiy the name of the dataset, preprocessing, tokenization and the training procedure in the config file to train an nlp model for that task.

install

pip install nmatheg

Configuration

Setup a config file for the training strategy.

[dataset]
dataset_name = ajgt_twitter_ar

[preprocessing]
segment = False
remove_special_chars = False
remove_english = False
normalize = False
remove_diacritics = False
excluded_chars = []
remove_tatweel = False
remove_html_elements = False
remove_links = False 
remove_twitter_meta = False
remove_long_words = False
remove_repeated_chars = False

[tokenization]
tokenizer_name = WordTokenizer
vocab_size = 1000
max_tokens = 128

[model]
model_name = rnn

[log]
print_every = 10

[train]
save_dir = .
epochs = 10
batch_size = 256

Main Sections

dataset describe the dataset and the task type. Currently we only support classification
preprocessing a set of cleaning functions mainly uses our library tnkeeh.
tokenization descrbies the tokenizer used for encoding the dataset. It uses our library tkseem.
train the training parameters like number of epochs and batch size.

Usage

Config Files

import nmatheg as nm
strategy = nm.TrainStrategy('config.ini')
strategy.start()

Benchmarking on multiple datasets and models

import nmatheg as nm
strategy = nm.TrainStrategy(
    datasets = 'arsentd_lev,arcd,caner', 
    models   = 'qarib/bert-base-qarib,aubmindlab/bert-base-arabertv01'
)
strategy.start()

Datasets

We are supporting huggingface datasets for Arabic. You can find the supported datasets here.

Dataset	Description
ajgt_twitter_ar	Arabic Jordanian General Tweets (AJGT) Corpus consisted of 1,800 tweets annotated as positive and negative. Modern Standard Arabic (MSA) or Jordanian dialect.
metrec	The dataset contains the verses and their corresponding meter classes. Meter classes are represented as numbers from 0 to 13. The dataset can be highly useful for further research in order to improve the field of Arabic poems’ meter classification. The train dataset contains 47,124 records and the test dataset contains 8,316 records.
labr	This dataset contains over 63,000 book reviews in Arabic. It is the largest sentiment analysis dataset for Arabic to-date. The book reviews were harvested from the website Goodreads during the month or March 2013. Each book review comes with the goodreads review id, the user id, the book id, the rating (1 to 5) and the text of the review.
ar_res_reviews	Dataset of 8364 restaurant reviews from qaym.com in Arabic for sentiment analysis
arsentd_lev	The Arabic Sentiment Twitter Dataset for Levantine dialect (ArSenTD-LEV) contains 4,000 tweets written in Arabic and equally retrieved from Jordan, Lebanon, Palestine and Syria.
oclar	The researchers of OCLAR Marwan et al. (2019), they gathered Arabic costumer reviews Zomato website on wide scope of domain, including restaurants, hotels, hospitals, local shops, etc. The corpus finally contains 3916 reviews in 5-rating scale. For this research purpose, the positive class considers rating stars from 5 to 3 of 3465 reviews, and the negative class is represented from values of 1 and 2 of about 451 texts.
emotone_ar	Dataset of 10,065 tweets in Arabic for Emotion detection in Arabic text
hard	This dataset contains 93,700 hotel reviews in Arabic language.The hotel reviews were collected from Booking.com website during June/July 2016.The reviews are expressed in Modern Standard Arabic as well as dialectal Arabic.The following table summarize some tatistics on the HARD Dataset.
caner	The Classical Arabic Named Entity Recognition corpus is a new corpus of tagged data that can be useful for handling the issues in recognition of Arabic named entities.
arcd	Arabic Reading Comprehension Dataset (ARCD) composed of 1,395 questions posed by crowdworkers on Wikipedia articles.

Tasks

Currently we support text classification, named entity recognition and question answering.

Demo

Check this colab notebook for a quick demo.

MagedSaeed/nmatheg