Ramitechs
  • Home
  • Services
  • Products
  • Careers
  • Contact

⇐ Back to Products

SentiArabic: A Sentiment Analysis Package for Standard Arabic


 The package includes:
- SLSA: A Sentiment Lexicon for Standard Arabic
- A Sentiment Corpus for Standard Arabic
- A Sentiment Analyzer for Standard Arabic 

  • Description
  • Publications
  • License
  • Prices
  • Request
<
>
SentiArabic is a complete sentiment analysis for Standard Arabic. The package include three main components:
- SLSA: A Sentiment Lexicon for Standard Arabic
- A Sentiment Corpus for Standard Arabic
- A Sentiment Analyzer for Standard Arabic

SLSA: A Sentiment Lexicon for Standard Arabic

SLSA is a state-of-the-art Sentiment Lexicon for Standard Arabic. The accuracy and richness of SLSA is proved to be superior to the other available lexicons (See the "Publication" tab). SLSA consists of 34,821 entries. An entry consists of:
     Arabic Lemma (in Buckwalter Transliteration, en.wikipedia.org/wiki/Buckwalter_transliteration), POS Tag (Noun, Adjective or Verb), Positive
     Score (0, 0.25, 0.5, 0.75 or 1), Negative Score (0, 0.25, 0.5, 0.75 or 1), Objective Score (1-Positive Score-Negative Score) and English Gloss.
SLSA currently has two versions:
     SLSA 1.0: An automatically generated version that is freely available for research purposes
     SLSA 2.0: The most frequent 5,000 entries are manually checked and the scores are better normalized
Examples:

Lemma POS Positive Negative Objective Gloss
munoSif_1 ADJ 0.5 0.0 0.5 equitable/fair
ta$A&um_1 NOUN 0.0 0.5 0.5 pessimism
waDaE VERB 0.0 0.0 0.0 put/place

A Sentiment Corpus for Standard Arabic

SentiArabic Corpus includes 20,000 sentences that are annotated for their contextual polarity (positive, negative or neutral).
The corpus covers a wide range of genres such as politics, sports, medicine, media, religions and fashion.
Examples:

Sentence Polarity
اعتادت أن تسحرنا بصوتها الملائكي الحالم Positive
لكنهم في وقت لاحق سيندمون Negative
النظام الغذائي يتدخل أيضاً في تحديد لون الأسنان Neutral

A Sentiment Analyzer for Standard Arabic

 SentiArabic sentiment analyzer is a state-of-the-art sentiment analyzer for Standard Arabic. The analyzer outperforms its counterparts by a significant margin when tested on blind test sets. The input to the analyzer is an Arabic sentence of a contextual polarity, and the output is the polarity of the sentence (positive or negative).
The analyzer can be accessed as an AWS service or obtained in full as part of the complete SentiArabic package.
Ramy Eskander and Owen Rambow
SLSA: A Sentiment Lexicon for Standard Arabic
In Proceedings of the 2015 Conference on Empirical Methods in Natural Language Processing (EMNLP)
Lisbon, Portugal, 2015

slsa.pdf
File Size: 211 kb
File Type: pdf
Download File

Ramy Eskander
SentiArabic: A Sentiment Analyzer for Standard Arabic
In Proceedings of the 2018 Conference on Language Resources and Evaluation (LREC)
Miyazaki, Japan, 2018

sentiarabic.pdf
File Size: 172 kb
File Type: pdf
Download File

# SentiArabic is not freely available, neither for research nor industrial purposes.
# You may not redistribute SentiArabic or any of its components.
# You may transform or rebuild SentiArabic or any of its components upon purchase.
# Exception: The free lexicon, SLSA 0.1, is freely available for research purposes.
# Publication: If you publish a paper on work that uses the lexicon, SLSA 0.1 or SLSA 0.2, appropriate credit should be given (see the “Publication” tab).
Component Research Purpose Non-research Purpose
SLSA Lexicon 1.0 $0 $840
SLSA Lexicon 2.0 $440 $1,215
SentiArabic Corpus 1.0 $345 $875
SentiArabic Analyzer Service 1.0 $160 per 100,000 requests $385 per 100,000 requests
Complete SentiArabic Package 1.0 $2,950 Contact Us
*All prices are in USD.
In order to request SentiArabic or any of its components, please send a message through the Contact page with the subject “SentiArabic”. You also need to include the following information:
     - Organization Name
     - Purpose of Use

  • Home
  • Services
  • Products
  • Careers
  • Contact