Elamur Research⌕
☰
home / papers / 2019 / Sentiment Analysis for Arabic in Social Media Network: A Systematic Mapping Study

Sentiment Analysis for Arabic in Social Media Network: A Systematic Mapping Study

Summary

The paper maps research on sentiment analysis of Arabic text from social networks (Arabic Sentiment Analysis, ASA). The authors note that earlier surveys chose papers without a repeatable procedure. They apply a systematic mapping protocol following Gough et al., guided by seven research questions. The questions cover techniques, machine-learning approaches, analysis level, research and contribution facets, data platforms, evaluation metrics and demographics. They queried Web of Science, ScienceDirect, IEEE Xplore and ACM for 2015–2019 work, then filtered the results in stages down to 51 primary studies. The authors report steady growth in ASA publications since 2015. Twitter is by far the dominant data source, and accuracy is the most common metric. Evaluation and solution research make up almost all studies, while validation work is rare. The authors argue that new ASA methods need more testing outside controlled settings. The paper is useful as an entry map for newcomers and as a template for transparent, countable literature selection on any NLP subfield.

Key points

  • Selection funnel: 277 hits (69 Web of Science, 147 ScienceDirect, 51 IEEE Xplore, 10 ACM) were narrowed to 51 studies in five steps: 236 → 204 → 146 → 97 → 51. The steps were the date window, de-duplication, removal of reviews and book chapters, title/abstract screening and full-text reading.
  • The search string paired "sentiment analysis" and "Social media" with Arabic, Arab or Arabizi. The paper distinguishes Modern Standard Arabic (MSA), Dialect Arabic (DA, typical of social media) and Classical Arabic (CA).
  • Research facets split as evaluation 49%, solution 47% and validation 4%. The abstract contradicts this by calling solution research the largest group.
  • Contribution facets include lexicon building (12 studies), models (10) and tools (11, about 22%), plus frameworks, approaches, methods and techniques. Doc2Vec and a BRST method are among the cited examples.
  • Data platforms: Twitter accounts for 60% (32 papers), Facebook 11%, YouTube 6% and TripAdvisor 4%. Others make up 19%, including Aljazeera, AL Arabia, Yahoo!-Maktoob and AWATIF.
  • The studies use 21 evaluation metrics in total. Accuracy leads at 29%, followed by precision (19%), recall (18%) and F-measure (16%).

Interesting ideas

  • I1. The selection pipeline reports how many items each filtering stage removed, so anyone can audit or repeat it. — For Mova (learn): Silver-data building could log how many sentences each filter drops, such as witness-parser disagreement or a teacher veto, making data loss visible and reproducible.
  • I2. Each work is tagged on two independent axes: the kind of research (solution, evaluation, validation) and the kind of artifact produced (lexicon, model, tool, framework). — For Mova (learn): Experiment records could carry both "what was tried" and "what artifact resulted", making it easy to see which components were only proposed and which were actually validated.
  • I3. The scarcity of validation research shows that many methods are judged only in controlled setups. This is a reminder to test systems beyond their home data. — For Mova (answer): Readers could be scored on held-out text genres (other fables, other school courses) as well as their tuning sets, so that results reflect real robustness.
  • I4. The heavy reliance on accuracy alone suggests that reporting several metrics gives a more honest picture, especially with skewed class balance. — For Mova (learn): Evaluation reports could always pair accuracy with precision, recall and F-measure, and compare against negative controls.
  • I5. The authors stress that informal dialect and romanized writing (Arabizi) differ from the formal standard, so data and tools must account for language variety. — For Mova (tokens): The tokenizer and tagger could mark text register (formal versus chat-like or nonstandard spelling) and route informal input through normalization before tagging.

Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.

Fields: Linguistics & languages

Cite

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.48550/arxiv.1911.05483 · arXiv 1911.05483

@misc{abo2019sentiment,
  title = {{Sentiment Analysis for Arabic in Social Media Network: A Systematic Mapping Study}},
  author = {Abo, Mohamed Elhag M. and Raj, Ram Gopal and Qazi, Atika and Zakari, Abubakar},
  year = {2019},
  doi = {10.48550/arxiv.1911.05483},
  eprint = {1911.05483},
  archivePrefix = {arXiv},
  url = {https://doi.org/10.48550/arxiv.1911.05483}
}

References

75 references, 2 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Main Findings: The main aim of this study is to investigate the current study in the ASA research area. In doing so, 51 primary studies were selected based on our methodology to be our primary studies for analysis. The following are the main findings of our study. a. The most frequently used in Arabic sentiment analysis is the Machine learning technique, which is used in 76% of the selected studies. However, we found growing activities in both lexicon-based and hybrid techniques. Another observation is the increasing number of different studies, with 24% in total out of all the primary studies. However, we observed that the Machine learning technique is still consistently gaining attention, whereby from (2015 – 2019), 76% of the studies are based on ML technique. Probably we would see more work on lexicon-based and hybrid technique in years to come because they got a promised result in the last few years. b. The supervised learning approach is the most frequently used in Arabic sentiment analysis is the Machine learning technique, which is used in 70% of the selected papers. However, unsupervised approach and semi-supervised approach are seen growing activities ASA research area. Furthermore, 18% of the selected studies used the unsupervised approach, while the semi-supervised approach has 12%. Moreover, the observation is the increasing number of another approach used in ASA studies. However, we observed that unsupervised learning approach is consistently gaining attention, in the last few years because of new machine learning algorithms not need training set such as Deep Learning. Probably we would see more work on unsupervised approach in years to come. c. In ASA research area, we found out that 75% of our selected primary studies are used word level in a supervised approach. Another observation is the slowly increasing number of miscellaneous studies, with 25% in total out of all the primary studies. Sentence level and document level showed 17%, and 8% respectively. d. On our research facet, 49% of the selected studies are evaluation papers, 47% are solution papers, and 4% are validation papers. Moreover, on contribution facet, Build Lexicon has 12 papers with 23%, followed by model and tools with 22%, the method with 12%, approach with 10%, framework four studies with 8%, and lastly, technique with 4%. e. We observed that datasets are the most collected form Twitter, which is 60% data in ASA field of research, which is 23 papers of the selected studies. Hence, the Twitter platform is the most utilized by the researcher. In general, social media networks dataset are the most utilized which have some of the smallest used of dataset among all the dataset platforms utilized in our primary studies. However, recent researchers are shifting to more realistic and streaming datasets with big data sentiment analysis. f. On our evaluation metrics, we observed that accuracy, recall, and precision expense score are the most utilized evaluation metrics by our selected studies. However, 9% of the evaluation metrics in primary selected studies were included as other evaluation metrics. Therefore, our analysis showed the top 10 evaluation metrics in the selected studies. g. The top publication fora's as identified from our primary papers are Journal of Procedia Computer Science and International Journal of Advanced Computer Science with 13 and six papers respectively. Information Processing and Management, Expert Systems with Applications, ACM Transactions on Asian and Low-Resource Language Information Processing (TALLIP), Egyptian Informatics Journal, Journal of King Saud University – Computer and Information Sciences, International Journal of Advanced and Applied Sciences, International Journal on Semantic Web and Information Systems, and Pattern Recognition Letters have made an immense contribution with 8 publications form top 10 journal. h. ASA research area has increased expanding consideration from the data science community since 2015, with an increasing number of publications with, on average, 45 publications from respectable journals and conferences every year. From our primary papers, we saw that about 59% of the research was published in journals, while 41% were distributed in conferences. With consistency in publication and increasing enthusiasm from the research community, we trust that the ASA area would presumably increase considerably more consideration the future. i. Because of the high activity in the ASA research area, 6% of the chose papers to have more than 20 citation counts. As far as nations with high publications, we have recognized that 29% of our primary papers were distributed from Saudi Arabia, while 13% of the studies were from Jordan. However, countries like the United States with Egypt (8%), Morocco 6%, and 5% for Algeria with Canada. Moreover, Jordan is published respectable papers out of 19 countries. Therefore, other institutions in the rest of the countries should do more effort to move the ASA research area forward in the near future. Despite Jordan being the second most active country in ASA research, the most influential author (Al-Kabi) is from a Zarqa University, Jordan, with four publications, followed by Al-Twairesh, Nora from the King Saud University, Riyadh, Saudi Arabia and Mohammad, Al-Ayyoub from Jordan University Science Technology, Jordan with 3 publications for each. We have noticed that of all our most influential writers, 45% are not from the most active countries (Saudi Arabia). find
  2. Moreover, the Arabic dialect dataset used for sentiment analysis, most of which was collected from Arab countries in Asia. Note that the Arab country in Africa represents 50% of the Arab countries. However, this dialect Arabic need more focus in future research. Considering the development of primary studies, we identified three main research type: evaluation, validation, and problem-solving. The research on the ASA seems to have high relevance on evaluation, which was evident in a large number of evaluation papers (49% of the primary studies) and a relatively decent number of problem-solving papers (47%). However, there was considerably less validation research (4%) in the literature. Therefore, this is a significant challenge in the ASA research area. Hence, for future work, more validation research is needed in the ASA research domain. Furthermore, some of the mentioned research challenges in the literature are also not sufficiently addressed despite much research effort. Research challenges such as Arabic dialect with lexicon-based because of lack of Arabic dialect dictionary[74], Building standard lexicon for Arabic dialect and modern standard Arabic [75] adopt a sentiment analysis methodology that works with different Arabic dialects[76]. To address these concerning issues, new standard dataset/techniques/approaches/ frameworks etc. need to largely address them. Perhaps the application of more deep learning techniques might aid to address these vital challenges. The ASA, through the application of machine learning, stands out when looking at methods and algorithms. Moreover, traditional sentiment analysis methods and algorithms, like SVM, NB, DT, and K-means, are frequently applied, and researches tend to enhance the text representation by applying NLP and deep learning methods or using large external datasets. Therefore, for future work, researchers are adopting DL and NLP for MSA and incorporated into the sentiment analysis process. Considering the development of primary studies, we identified three main ASA levels: document level, sentence level, and word or aspect level. We expect an increase in the number of studies that have document level because it showed only four studies under a supervised approach, which is 8%. We also expect a raise of resources (linguistic resources and annotated corpora) for Arabic languages. These resources are essential for the development of ASA technologies. The greater availability of Arabic resources will provide more studies on these languages. 5. Conclusion The Arabic language started takes place in sentiment analysis research since it has a significant impact on text meaning. Be that as it may, there is an absence of secondary studies that consolidate these researchers. This study an SMS directed to a review of Arabic concerned sentiment analysis literature. The scope of this study is broad for the first search (277 papers matched the search expression). Therefore, the mapping was chiefly performed dependent on abstracts and critical reading of papers. By and by, we trust that our limitations do crucially affect the outcomes since our research has wide coverage. The principal contributions of our study are (1) it exhibits a quantitative analysis of the research ASA; (2) its pursued a well-characterized convention for SMS; (3) it discusses the area of ASA in regards to seven crucial aspects: the technique, machine learning approaches, level of sentiment analysis, Research facets, contribution facets, applications platforms, and demographic characteristics); and (4) the created mapping can contribute an overall outline of the ASA and can be of extraordinary assistance for conducting researchers in sentiment analysis and text mining. In this way, this effort can fill a gap in the literature as supposedly, to the best of our knowledge, this is the first general systematic mapping study of this extensive topic for ASA. Although a few kinds of research have been developed in the sentiment analysis field the processing of Arabic text remains an open research issue. The field needs secondary studies in areas that have a few numbers of primary studies, such as lexicon-based and hybrid approach. An extra point in Arabic language concern. We found extensive differences in quantities of concentrates among modern standard Arabic and Arabic dialect since 56% of the recognized studies manage MSA. In this manner, there is an absence of studies managing dealing with texts written in Arabic dialect. When considering Arabic dialect text mining, we trust that this need can be filled up with the development of good knowledge bases and natural language processing methods specific to these languages. In addition, the sentiment analysis effect of dialects languages is additionally a motivating open research question in light of the variance of Arabic dialect. A correlation among feature aspects of various dialects and their effect on the results of sentiment analysis methods would be interesting. References 1. Proksch, S.O., et al., Multilingual sentiment analysis: A new approach to measuring conflict in legislative speeches. Legislative Studies Quarterly, 2019. 44(1): p. 97-131. 2. Abo, M.E.M., et al. SSA-SDA: Subjectivity and Sentiment Analysis of Sudanese Dialect Arabic. in 2019 International Conference on Computer and Information Sciences (ICCIS). 2019. IEEE. 3. Al-Moslmi, T., et al., Arabic senti-lexicon: Constructing publicly available language resources for Arabic sentiment analysis. Journal of Information Science, 2018. 44(3): p. 345-362. find
  3. Al-Ayyoub, M., et al., A comprehensive survey of arabic sentiment analysis. Information Processing & Management, 2019. 56(2): p. 320-342. find
  4. Al-Twairesh, N., et al., SUAR: Towards Building a Corpus for the Saudi Dialect. Procedia Computer Science, 2018. 142: p. 72-82. find
  5. Almanie, T., et al. Saudi Mood: A Real-Time Informative Tool for Visualizing Emotions in Saudi Arabia Using Twitter. in 2018 21st Saudi Computer Society National Computer Conference (NCC). 2018. find
  6. Alahmary, R.M., H.Z. Al-Dossari, and A.Z. Emam. Sentiment Analysis of Saudi Dialect Using Deep Learning Techniques. in 2019 International Conference on Electronics, Information, and Communication (ICEIC). 2019. find
  7. Baly, R., et al., Comparative Evaluation of Sentiment Analysis Methods Across Arabic Dialects. Procedia Computer Science, 2017. 117: p. 266-273. find
  8. Alnawas, A. and N. Arıcı, The corpus based approach to sentiment analysis in modern standard Arabic and Arabic dialects: A literature review. Politeknik Dergisi, 2018. 21(2): p. 461-470. find
  9. Boudad, N., et al., Sentiment analysis in Arabic: A review of the literature. Ain Shams Engineering Journal, 2017. find
  10. Salloum, S.A., et al., A survey of Arabic text mining, in Intelligent Natural Language Processing: Trends and Applications. 2018, Springer. p. 417-431. find
  11. El-Masri, M., N. Altrabsheh, and H. Mansour, Successes and challenges of Arabic sentiment analysis research: a literature review. Social Network Analysis and Mining, 2017. 7(1): p. 54. find
  12. Biltawi, M., et al. Sentiment classification techniques for Arabic language: A survey. in 2016 7th International Conference on Information and Communication Systems (ICICS). 2016. IEEE. find
  13. Sinoara, R.A., J. Antunes, and S.O. Rezende, Text mining and semantics: a systematic mapping study. Journal of the Brazilian Computer Society, 2017. 23(1): p. 9. find
  14. Gough, D., S. Oliver, and J. Thomas, An introduction to systematic reviews. 2017: Sage. find
  15. Al-Smadi, M., et al., Enhancing aspect-based sentiment analysis of Arabic hotels’ reviews using morphological, syntactic and semantic features. Information Processing & Management, 2019. 56(2): p. 308-319. find
  16. Heikal, M., M. Torki, and N. El-Makky, Sentiment analysis of Arabic Tweets using deep learning. Procedia Computer Science, 2018. 142: p. 114-122. find
  17. AlMurtadha, Y., Mining Trending Hash Tags for Arabic Sentiment Analysis. International Journal of Advanced Computer Science and Applications, 2018. 9(2): p. 189-194. find
  18. Al-Smadi, M., et al., Deep Recurrent neural network vs. support vector machine for aspect-based sentiment analysis of Arabic hotels’ reviews. Journal of computational science, 2018. 27: p. 386-393. find
  19. Alnawas, A. and N. Arici, Sentiment Analysis of Iraqi Arabic Dialect on Facebook Based on Distributed Representations of Documents. ACM Trans. Asian Low-Resour. Lang. Inf. Process., 2019. 18(3): p. 1-17. find
  20. Atoum, J.O. and M. Nouman, Sentiment Analysis of Arabic Jordanian Dialect Tweets. International Journal of Advanced Computer Science and Applications, 2019. 10(2): p. 256-262. find
  21. Lulu, L. and A. Elnagar, Automatic Arabic Dialect Classification Using Deep Learning Models. Procedia Computer Science, 2018. 142: p. 262-269. find
  22. Youssef, M. and S.R. El-Beltagy, MoArLex: An Arabic Sentiment Lexicon Built Through Automatic Lexicon Expansion. Procedia Computer Science, 2018. 142: p. 94-103. find
  23. Bin Hathlian, N.F. and A.M. Hafez, Subjective Text Mining for Arabic Social Media. International Journal on Semantic Web and Information Systems, 2017. 13(2): p. 1-13. find
  24. Tellez, E.S., et al., A simple approach to multilingual polarity classification in Twitter. Pattern Recognition Letters, 2017. 94: p. 68-74. find
  25. Abo, M.E.M., et al. Sentiment analysis algorithms: evaluation performance of the Arabic and English language. in 2018 International Conference on Computer, Control, Electrical, and Electronics Engineering (ICCCEEE). 2018. find
  26. Chen, X., et al., Adversarial deep averaging networks for cross-lingual sentiment classification. Transactions of the Association for Computational Linguistics, 2018. 6: p. 557-570. find our card
  27. Tubishat, M., et al., Improved whale optimization algorithm for feature selection in Arabic sentiment analysis. Applied Intelligence, 2019. 49(5): p. 1688-1707. find
  28. Oussous, A., et al., ASA: A framework for Arabic sentiment analysis. Journal of Information Science, 2019: p. 0165551519849516. find
  29. Al-Sallab, A., et al., Aroma: A recursive deep learning model for opinion mining in arabic as a low resource language. ACM Transactions on Asian and Low-Resource Language Information Processing (TALLIP), 2017. 16(4): p. 25. find
  30. Abd-Elhamid, L., D. Elzanfaly, and A.S. Eldin. Feature-based sentiment analysis in online Arabic reviews. in 2016 11th International Conference on Computer Engineering & Systems (ICCES). 2016. IEEE. find
  31. Al-Twairesh, N., et al., Sentiment analysis of arabic tweets: Feature engineering and a hybrid approach. arXiv preprint arXiv:1805.08533, 2018. ↗
  32. Baly, R., et al., ArSentD-LEV: A multi-topic corpus for target-based sentiment analysis in Arabic levantine tweets. arXiv preprint arXiv:1906.01830, 2019. ↗ our card
  33. Al-Azani, S. and E.-S. El-Alfy. Emojis-based sentiment classification of Arabic microblogs using deep recurrent neural networks. in 2018 International Conference on Computing Sciences and Engineering (ICCSE). 2018. IEEE. find
  34. Al-Twairesh, N., et al., AraSenTi-Tweet: A Corpus for Arabic Sentiment Analysis of Saudi Tweets. Procedia Computer Science, 2017. 117: p. 63-72. find
  35. Touahri, I. and A. Mazroui, Studying the effect of characteristic vector alteration on Arabic sentiment classification. Journal of King Saud University - Computer and Information Sciences, 2019. find
  36. Guellil, I. and F. Azouaou. Arabic Dialect Identification with an Unsupervised Learning (Based on a Lexicon). Application Case: ALGERIAN Dialect. in 2016 IEEE Intl Conference on Computational Science and Engineering (CSE) and IEEE Intl Conference on Embedded and Ubiquitous Computing (EUC) and 15th Intl Symposium on Distributed Computing and Applications for Business Engineering (DCABES). 2016. find
  37. Duwairi, R.M., N.A. Ahmed, and S.Y. Al-Rifai, Detecting sentiment embedded in Arabic social media - A lexicon-based approach. Journal of Intelligent & Fuzzy Systems, 2015. 29(1): p. 107-117. find
  38. Al-Thubaity, A., Q. Alqahtani, and A. Aljandal, Sentiment lexicon for sentiment analysis of Saudi dialect tweets. Procedia Computer Science, 2018. 142: p. 301-307. find
  39. Alami, N., M. Meknassi, and N. En-nahnahi, Enhancing unsupervised neural networks based text summarization with word embedding and ensemble learning. Expert Systems with Applications, 2019. 123: p. 195-211. find
  40. Bekkali, M. and A. Lachkar, Arabic Sentiment Analysis based on Topic Modeling, in Proceedings of the New Challenges in Data Sciences: Acts of the Second Conference of the Moroccan Classification Society. 2019, ACM: Kenitra, Morocco. p. 1-6. find
  41. Elnagar, A., L. Lulu, and O. Einea, An Annotated Huge Dataset for Standard and Colloquial Arabic Reviews for Subjective Sentiment Analysis. Procedia Computer Science, 2018. 142: p. 182-189. find
  42. Abdul-Mageed, M., Modeling Arabic subjectivity and sentiment in lexical space. Information Processing & Management, 2019. 56(2): p. 291-307. find
  43. Elshakankery, K. and M.F. Ahmed, HILATSA: A hybrid Incremental learning approach for Arabic tweets sentiment analysis. Egyptian Informatics Journal, 2019. find
  44. Jardaneh, G., et al. Classifying Arabic Tweets Based on Credibility Using Content and User Features. in 2019 IEEE Jordan International Joint Conference on Electrical Engineering and Information Technology (JEEIT). 2019. find
  45. Al-Azani, S. and E. El-Alfy. Emojis-Based Sentiment Classification of Arabic Microblogs Using Deep Recurrent Neural Networks. in 2018 International Conference on Computing Sciences and Engineering (ICCSE). 2018. find
  46. Al–Azani, S. and E.M. El–Alfy. Imbalanced Sentiment Polarity Detection Using Emoji-Based Features and Bagging Ensemble. in 2018 1st International Conference on Computer Applications & Information Security (ICCAIS). 2018. find
  47. Soumeur, A., et al., Sentiment Analysis of Users on Social Networks: Overcoming the challenge of the Loose Usages of the Algerian Dialect. Procedia Computer Science, 2018. 142: p. 26-37. find
  48. Alhamid, M.F., et al. Detection and Visualization of Arabic Emotions on Social Emotion Map. in 2017 IEEE International Symposium on Multimedia (ISM). 2017. find
  49. Al-Hussaini, H. and H. Al-Dossari, A Lexicon-based Approach to Build Service Provider Reputation from Arabic Tweets in Twitter. International Journal of Advanced Computer Science and Applications, 2017. 8(4): p. 445-454. find
  50. Al-Shabi, A., et al., Cross-Lingual Sentiment Classification from English to Arabic using Machine Translation. International Journal of Advanced Computer Science and Applications, 2017. 8(12): p. 434-440. find
  51. Altaher, A., Hybrid approach for sentiment analysis of Arabic tweets based on deep learning model and features weighting. International Journal of Advanced and Applied Sciences, 2017. 4(8): p. 43-49. find
  52. Alwakid, G., T. Osman, and T. Hughes-Roberts, Challenges in Sentiment Analysis for Arabic Social Networks. Procedia Computer Science, 2017. 117: p. 89-100. find
  53. El-Naggar, N., Y. El-Sonbaty, and M.A. El-Nasr. Sentiment analysis of modern standard Arabic and Egyptian dialectal Arabic tweets. in 2017 Computing Conference. 2017. find
  54. El-Masri, M., et al., A web-based tool for Arabic sentiment analysis. Procedia Computer Science, 2017. 117: p. 38-45. find
  55. Mallek, F., B. Belainine, and F. Sadat, Arabic Social Media Analysis and Translation. Procedia Computer Science, 2017. 117: p. 298-303. find
  56. Al-Kabi, M.N., H.A. Wahsheh, and I.M. Alsmadi, Polarity Classification of Arabic Sentiments. International Journal of Information Technology and Web Engineering, 2016. 11(3): p. 32-49. find
  57. Bouchlaghem, R., A. Elkhelifi, and R. Faiz, A Machine Learning Approach For Classifying Sentiments in Arabic tweets, in Proceedings of the 6th International Conference on Web Intelligence, Mining and Semantics. 2016, ACM: Nîmes, France. p. 1-6. find
  58. Hathlian, N.F.B. and A.M. Hafezs. Sentiment - subjective analysis framework for arabic social media posts. in 2016 4th Saudi International Conference on Information Technology (Big Data Analysis) (KACSTIT). 2016. find
  59. Hussien, W.A., et al. Are emoticons good enough to train emotion classifiers of Arabic tweets? in 2016 7th International Conference on Computer Science and Information Technology (CSIT). 2016. find
  60. Mohammad, S.M., M. Salameh, and S. Kiritchenko, How Translation Alters Sentiment. Journal of Artificial Intelligence Research, 2016. 55: p. 95-130. find
  61. Rabab'ah, A.M., et al. Evaluating SentiStrength for Arabic Sentiment Analysis. in 2016 7th International Conference on Computer Science and Information Technology (CSIT). 2016. find
  62. AL-Smadi, M., et al., Using aspect-based sentiment analysis to evaluate arabic news affect on readers, in Proceedings of the 8th International Conference on Utility and Cloud Computing. 2015, IEEE Press: Limassol, Cyprus. p. 436-441. find
  63. Cherif, W., A. Madani, and M. Kissi, Towards an Efficient Opinion Measurement in Arabic Comments. Procedia Computer Science, 2015. 73: p. 122-129. find
  64. Alwehaibi, A. and K. Roy. Comparison of Pre-Trained Word Vectors for Arabic Text Classification Using Deep Learning Approach. in 2018 17th IEEE International Conference on Machine Learning and Applications (ICMLA). 2018. find
  65. AlOtaibi, S. and M.B. Khan, Discovering Semantic and Sentiment Correlations using Short Informal Arabic Language Text. International Journal of Advanced Computer Science and Applications, 2017. 8(1): p. 198-207. find
  66. Khasawneh, R.T., et al. Arabic sentiment polarity identification using a hybrid approach. in 2015 6th International Conference on Information and Communication Systems (ICICS). 2015. find
  67. Wang, H., et al. A System for Extracting Sentiment from Large-Scale Arabic Social Data. in 2015 First International Conference on Arabic Computational Linguistics (ACLing). 2015. find
  68. Mataoui, M., et al. A new syntax-based aspect detection approach for sentiment analysis in Arabic reviews. in 2018 2nd International Conference on Natural Language and Speech Processing (ICNLSP). 2018. find
  69. El-Halees, A.M. Arabic Opinion Mining Using Distributed Representations of Documents. in 2017 Palestinian International Conference on Information and Communication Technology (PICICT). 2017. find
  70. Al-Twairesh, N., H. Al-Khalifa, and A. Al-Salman. Towards Analyzing Saudi Tweets. in 2015 First International Conference on Arabic Computational Linguistics (ACLing). 2015. find
  71. Joshi, A., Sentiment Analysis and Opinion Mining from Noisy Social Media Content. 2019, International Institute of Information Technology Hyderabad. find
  72. AlOtaibi, S. and M.B. Khan, Sentiment Analysis Challenges of Informal Arabic Language. International Journal of Advanced Computer Science and Applications, 2017. 8(2): p. 278-284. find
  73. Alshutayri, A. and E. Atwell. Creating an Arabic Dialect Text Corpus by Exploring Twitter, Facebook, and Online Newspapers. in Proceedings of OSACT'2018 Open-Source Arabic Corpora and Processing Tools. 2018. Leeds. find
  74. Jarrar, M. and H. Amayreh. An Arabic-Multilingual Database with a Lexicographic Search Engine. in International Conference on Applications of Natural Language to Information Systems. 2019. Springer. find
  75. Freitas, P., et al. Social Web Mining as a Tool to Support Public Security Sentiment Analysis. in EmC-ICDSST 2019 5 th International Conference on Decision Support System Technology–ICDSST 2019 & EURO Mini Conference 2019 on “Decision Support Systems: Main. 2019. find

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/abo-2019-sentiment-analysis-arabic-social-media/